JSDM models test
Data description
Data was simulated using VirtualCommunity code.
Simulated data contains 20 data sets.
NB - if chain/sample parameters is modified - should be changed in the whole section(dataset)
Environment filtering 5 species
JSDM
data<-sim_data$EnvEvenSp5
data <- list(
Y = subset(data, select = -env),
X = cbind(1, scale(poly(data$env, 2))),
covx = cov(cbind(1, scale(poly(data$env, 2)))),
K = 3,
J = ncol(data) - 1,
n = nrow(data),
I = diag(ncol(data) - 1),
df = ncol(data)
)
Y_cor<-cor(data$Y)
to_prec<-function(m){
n<-dim(m)[1]
Tau_n<-matrix(nrow=n, ncol=n)
for (j in 1:n) {
for (k in 1:n){
Tau_n[j, k] <- -m[j, k]/sqrt((m[j,j]*m[k,k]))
}
}
return(Tau_n)
}
#Tau_n<-matrix(nrow=dim(model$mean$Tau)[1], ncol=dim(model$mean$Tau)[1])
Tau_n<-to_prec(me5$mean$Tau)
#Tau_k<-Tau_n*(!(model$q97.5$Tau>0 & model$q2.5$Tau<0))
par(mfrow=c(2,4),oma = c(3, 1, 2, 1))
cols = colorRampPalette(c("dark blue","white","red"))
col2 <- colorRampPalette(c("#4393C3", "#2166AC", "#053061",
"#FDDBC7", "#FFFFFF", "#D1E5F0", "#92C5DE",
"#67001F", "#B2182B", "#D6604D", "#F4A582"))
corrplot(Y_cor, diag = FALSE, order = "original",tl.pos = "ld", tl.cex = 0.5, method = "color",col=cols(200), type = "lower")
title("Correlation cor(Y)")
corrplot(me5$mean$EnvRho, diag = FALSE, order = "original",tl.pos = "ld", tl.cex = 0.5, method = "color",col=cols(200), type = "lower")
title("EnvRho")
corrplot(me5$mean$EnvRho*(!me5$overlap0$EnvRho), diag = FALSE, order = "original",tl.pos = "ld", tl.cex = 0.5, method = "color",col=cols(200), type = "lower")
title("EnvRho signif")
corrplot(me5$mean$Rho, diag = FALSE, order = "original",tl.pos = "ld", tl.cex = 0.5, method = "color",col=cols(200), type = "lower")
title("Rho")
corrplot(me5$mean$Rho*(!me5$overlap0$Rho), diag = FALSE, order = "original",tl.pos = "ld", tl.cex = 0.5, method = "color",col=cols(200), type = "lower")
title("Rho signif")
corrplot(Tau_n, diag = FALSE, order ="original",tl.pos = "ld", tl.cex = 0.5, method = "color",col=cols(200), type = "lower")
title("Tau")
corrplot(Tau_n*(!me5$overlap0$Tau), diag = FALSE, order ="original",tl.pos = "ld", tl.cex = 0.5, method = "color",col=cols(200), type = "lower")
title("Tau signif")
GJAM
##to setup chains parameters
data<-sim_data$EnvEvenSp5
#fit_gjam(data,2000,1000,"./gjam_models/gjam5env.rda")
load_gjam(data,name="./gjam_models/gjam5env.rda")##
## Sensitivity by predictor variables f:
## Estimate SE CI_025 CI_975
## env 9.5 3.13 5.08 16.4
## env2 31.9 7.93 20.50 50.6
##
## Coefficient matrix B:
## sp01 sp02 sp03 sp04 sp05
## intercept -0.0312 0.646 1.2500 0.704 -1.200
## env -2.4200 -1.800 0.0438 1.710 2.870
## env2 0.9440 -0.527 -0.9410 -0.598 -0.665
## RMSPE 0.2780 0.327 0.3760 0.309 0.327
##
## Coefficient matrix B:
## Estimate SE CI_025 CI_975 sig95
## sp01_intercept -0.0312 0.0775 -0.1940 0.107
## sp01_env -2.4200 0.1170 -2.6500 -2.200 *
## sp01_env2 0.9440 0.0938 0.7780 1.130 *
## sp02_intercept 0.6460 0.0651 0.5220 0.771 *
## sp02_env -1.8000 0.0911 -1.9900 -1.630 *
## sp02_env2 -0.5270 0.0741 -0.6700 -0.389 *
## sp03_intercept 1.2500 0.0672 1.1100 1.380 *
## sp03_env 0.0438 0.0465 -0.0447 0.132
## sp03_env2 -0.9410 0.0764 -1.0900 -0.781 *
## sp04_intercept 0.7040 0.0865 0.5590 0.881 *
## sp04_env 1.7100 0.0897 1.5500 1.890 *
## sp04_env2 -0.5980 0.0608 -0.7170 -0.486 *
## sp05_intercept -1.2000 0.0953 -1.3700 -1.000 *
## sp05_env 2.8700 0.1380 2.6300 3.170 *
## sp05_env2 -0.6650 0.0610 -0.7840 -0.545 *
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Coefficient matrix B, standardized for X:
## NULL
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Coefficient matrix B, standardized for X and W:
## NULL
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Design Table
## env env2
## VIF 1 1
## env2 0 NA
##
## Sample contains n = 500 observations on S = 5 response variables, and 2 predictors. Data types (typeNames) include PA. There are 0 missing values in X and 0 missing values in Y. The RMSPE is 0.325, and the DIC is 25294. Computation involved 2000 Gibbs steps, with a burnin of 1000.
HMSC
#setwd("~/Tesi/Code/Ecology-models-master/simcoms-master/ExampleFiles")
fit_hmsc<-function(data,label="Fit",nsamples = 1000,nchains=2,name="./HMmodels/hmtemp.rda" ){
if (label=="Fit"){
Y_data = subset(data, select = -env)
ns<- ncol(Y_data)
np <- nrow(Y_data)
X<-scale(poly(data$env[1:np], 2))
colnames(X)<-c("env","env2")
studyDesign = data.frame(sample = as.factor(1:np))
rL = HmscRandomLevel(units = studyDesign$sample)
m = Hmsc(Y=as.matrix(Y_data), XData=as.data.frame(X), XFormula=~env+env2, distr="probit",
studyDesign = studyDesign, ranLevels = list(sample = rL))
m = sampleMcmc(m, nsamples, thin=10, adaptNf=c(200,200), transient=500,nChains=nchains ,verbose=F)
save(m, file = name)
return(m)
}
if (label=="Load"){
return(load_object(name))
}
}
data<-sim_data$EnvEvenSp5
hm_mod<-fit_hmsc(data,"Load",name="./HMmodels/hm5env.rda" )
#hm_mod<-load_object("./HMmodels/hm5env.rda")Convergence:
hm_conv<-function(mod){
codaList = convertToCodaObject(mod)
#convergence histograms
hist(effectiveSize(codaList$Beta), main="ess(beta)",lwd=2,col=gray(.6))
hist(gelman.diag(codaList$Beta,multivariate=FALSE)$psrf,lwd=2,col=gray(.6), main="psrf(beta)")
hist(effectiveSize(codaList$Omega[[1]]), main="ess(omega)",lwd=2,col=gray(.6))
hist(gelman.diag(codaList$Omega[[1]], multivariate=FALSE)$psrf,lwd=2,col=gray(.6), main="psrf(omega)")
}
hm_conv(hm_mod)
Study of interactions
hm_inter<-function(mod, nchains=2,nsamples = 1000, interact=diag(ns)){
getOmega = function(a,r=1)
return(crossprod(a$Lambda[[r]]))
ns<-mod$ns
postOmega1 = array(unlist(lapply(mod$postList[[1]],getOmega)),c(ns,ns,mod$samples))
postOmega2 = array(unlist(lapply(mod$postList[[2]],getOmega)),c(ns,ns,mod$samples))
postOmega<-abind(postOmega1,postOmega2,along=3)
postOmegaMean = apply(postOmega,c(1,2),mean)
postOmegaUp=apply(postOmega,c(1,2),quantile,0.95)
postOmegaLo=apply(postOmega,c(1,2),quantile,0.05)
postR<-array(dim=c(ns,ns,nchains*nsamples))
for(i in 1:dim(postOmega)[3])
postR[,,i]<-stats::cov2cor(postOmega[,,i])
postRMean = apply(postR,c(1,2),mean)
postRUp=apply(postR,c(1,2),quantile,0.95)
postRLo=apply(postR,c(1,2),quantile,0.05)
Tau = solve(postOmegaMean)
Tau_n = -cov2cor(Tau)
Toplot_R<-postRMean*(!(postRUp>0 & postRLo<0))
# Omegacor<- computeAssociations(m)
# supportLevel<- 0.95
# toPlot<- ((Omegacor[[1]]$support>supportLevel)+ (Omegacor[[1]]$support<(1-supportLevel))>0)*Omegacor[[1]]$mean
# corrplot(toPlot, method="color", col=colorRampPalette(c("blue", "white", "red"))(200))
par(mfrow=c(2,3),oma = c(1, 1, 1, 1))
corrplot(cor(hm_mod$Y), diag = FALSE, order = "original",tl.pos = "ld", tl.cex = 0.5, method = "color",col=cols(200), type = "lower")
title("Correlation cor(Y)")
corrplot(postRMean, diag = FALSE, order = "original",tl.pos = "ld", tl.cex = 0.5, method = "color",col=cols(200), type = "lower")
title("R")
corrplot(Toplot_R, diag = FALSE, order = "original",tl.pos = "ld", tl.cex = 0.5, method = "color",col=cols(200), type = "lower")
title("Plot only non zero value")
corrplot(Tau_n, diag = FALSE, order = "original",tl.pos = "ld", tl.cex = 0.5, method = "color",col=cols(200), type = "lower")
title("Partial correlation matrix")
corrplot(interact, diag = FALSE, order = "original",tl.pos = "ld", tl.cex = 0.5, method = "color",col=cols(200), type = "lower")
title("True interactions")
}
hm_inter(hm_mod)
Environmental filtering 10 species
JSDM
me10 <- load_object("model-2019-04-10-08-26-20.rda")
#load("model-2019-04-09-19-02-16.rda")
summary(me10)## Summary for model '/tmp/RtmpKix4lZ/file40e957831178'
## Saved parameters: B Rho EnvRho
## MCMC ran in parallel for 683.734 minutes at time 2019-04-09 21:02:35.
##
## For each of 5 chains:
## Adaptation: 250000 iterations (possibly insufficient)
## Burn-in: 0 iterations
## Thin rate: 100 iterations
## Total chain length: 270000 iterations
## Posterior sample size: 200 draws
##
## **WARNING** Rhat values indicate convergence failure.jsdm_conv(me10)## Maximum Rhat value for Beta: 1.5602
## Maximum Rhat value for Rho: NA
## Maximum Rhat value for EnvRho: 1.7953
me10$mcmc.info[1:7]## $n.chains
## [1] 5
##
## $n.adapt
## [1] 250000 250000 250000 250000 250000
##
## $sufficient.adapt
## [1] FALSE FALSE FALSE FALSE FALSE
##
## $n.iter
## [1] 20000
##
## $n.burnin
## [1] 0
##
## $n.thin
## [1] 100
##
## $n.samples
## [1] 1000data<-sim_data$EnvEvenSp10
data <- list(
Y = subset(data, select = -env),
X = cbind(1, scale(poly(data$env, 2))),
covx = cov(cbind(1, scale(poly(data$env, 2)))),
K = 3,
J = ncol(data) - 1,
n = nrow(data),
I = diag(ncol(data) - 1),
df = ncol(data)
)
##########################################################################################
#Tau<-solve
#Tau_n<-to_prec(me10$mean$Tau)
#Tau_k<-Tau_n*(!(model$q97.5$Tau>0 & model$q2.5$Tau<0))
plot_cor_jsdm<-function(mod,y,interact=diag(ncol(y))){
par(mfrow=c(2,4),oma = c(3, 1, 2, 1))
corrplot(cor(y), diag = FALSE, order = "original",tl.pos = "ld", tl.cex = 0.5, method = "color",col=cols(200), type = "lower")
title("Correlation cor(Y)")
corrplot(mod$mean$EnvRho, diag = FALSE, order = "original",tl.pos = "ld", tl.cex = 0.5, method = "color",col=cols(200), type = "lower")
title("EnvRho")
corrplot(mod$mean$EnvRho*(!mod$overlap0$EnvRho), diag = FALSE, order = "original",tl.pos = "ld", tl.cex = 0.5, method = "color",col=cols(200), type = "lower")
title("EnvRho signif")
corrplot(mod$mean$Rho, diag = FALSE, order = "original",tl.pos = "ld", tl.cex = 0.5, method = "color",col=cols(200), type = "lower")
title("Rho")
corrplot(mod$mean$Rho*(!mod$overlap0$Rho), diag = FALSE, order = "original",tl.pos = "ld", tl.cex = 0.5, method = "color",col=cols(200), type = "lower")
title("Rho signif")
corrplot(interact, diag = FALSE, order = "original",tl.pos = "ld", tl.cex = 0.5, method = "color",col=cols(200), type = "lower")
title("True interactions")
}
plot_cor_jsdm(me10,data$Y)
Gjam
data<-sim_data$EnvEvenSp10
#fit_gjam(data,5000,500,"./gjam_models/gjam10env.rda")
load_gjam(data,name="./gjam_models/gjam10env.rda")##
## Sensitivity by predictor variables f:
## Estimate SE CI_025 CI_975
## env 233.0 58.0 120 354.0
## env2 56.7 11.8 37 82.9
##
## Coefficient matrix B:
## sp01 sp02 sp03 sp04 sp05 sp06 sp07 sp08 sp09
## intercept -0.2060 0.3800 1.310 0.889 1.020 0.742 0.638 0.140 0.332
## env -2.0500 -2.5700 -2.300 -1.030 -0.189 0.542 1.090 1.740 2.520
## env2 0.0656 -0.0531 0.338 -0.533 -0.805 -0.838 -0.690 -0.425 0.640
## RMSPE 0.3300 0.2980 0.317 0.371 0.414 0.421 0.374 0.345 0.298
## sp10
## intercept -0.577
## env 2.370
## env2 0.677
## RMSPE 0.276
##
## Coefficient matrix B:
## Estimate SE CI_025 CI_975 sig95
## sp01_intercept -0.2060 0.0670 -0.33700 -0.0733 *
## sp01_env -2.0500 0.1190 -2.29000 -1.8300 *
## sp01_env2 0.0656 0.0992 -0.14100 0.2380
## sp02_intercept 0.3800 0.1030 0.18000 0.5510 *
## sp02_env -2.5700 0.1370 -2.83000 -2.2900 *
## sp02_env2 -0.0531 0.0790 -0.19700 0.1120
## sp03_intercept 1.3100 0.0829 1.15000 1.4700 *
## sp03_env -2.3000 0.1260 -2.55000 -2.0600 *
## sp03_env2 0.3380 0.0712 0.20500 0.4860 *
## sp04_intercept 0.8890 0.0585 0.77400 1.0000 *
## sp04_env -1.0300 0.0659 -1.16000 -0.9070 *
## sp04_env2 -0.5330 0.0580 -0.64200 -0.4160 *
## sp05_intercept 1.0200 0.0621 0.90800 1.1500 *
## sp05_env -0.1890 0.0488 -0.28500 -0.0949 *
## sp05_env2 -0.8050 0.0759 -0.96000 -0.6590 *
## sp06_intercept 0.7420 0.0635 0.61600 0.8650 *
## sp06_env 0.5420 0.0558 0.43200 0.6460 *
## sp06_env2 -0.8380 0.0634 -0.96100 -0.7130 *
## sp07_intercept 0.6380 0.0795 0.50600 0.8160 *
## sp07_env 1.0900 0.0805 0.93900 1.2500 *
## sp07_env2 -0.6900 0.0595 -0.80300 -0.5700 *
## sp08_intercept 0.1400 0.0759 0.00798 0.2950 *
## sp08_env 1.7400 0.1240 1.48000 1.9600 *
## sp08_env2 -0.4250 0.1000 -0.65400 -0.2750 *
## sp09_intercept 0.3320 0.0539 0.22600 0.4370 *
## sp09_env 2.5200 0.1660 2.22000 2.8500 *
## sp09_env2 0.6400 0.0700 0.50400 0.7780 *
## sp10_intercept -0.5770 0.0696 -0.70200 -0.4380 *
## sp10_env 2.3700 0.1130 2.16000 2.6000 *
## sp10_env2 0.6770 0.0820 0.49900 0.8200 *
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Coefficient matrix B, standardized for X:
## NULL
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Coefficient matrix B, standardized for X and W:
## NULL
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Design Table
## env env2
## VIF 1 1
## env2 0 NA
##
## Sample contains n = 500 observations on S = 10 response variables, and 2 predictors. Data types (typeNames) include PA. There are 0 missing values in X and 0 missing values in Y. The RMSPE is 0.348, and the DIC is 114934. Computation involved 5000 Gibbs steps, with a burnin of 500.
#gje10<-load_object("./gjam_models/gjam10env.rda")
#to check posterior density of s in Sigma
#gje10<-load_object("./gjam_models/gjam10env.rda")
#plot(density(gje10$chains$sgibbs[,4]))HMSC
data<-sim_data$EnvEvenSp10
hm_mod<-fit_hmsc(data,"Load",name="./HMmodels/hm10env.rda" )
hm_conv(hm_mod)
hm_inter(hm_mod)
Environmental filtering 20 species
JSDM
me20 <- load_object("model-2019-04-11-19-06-02.rda")
#load("model-2019-04-09-19-02-16.rda")
summary(me20)## Summary for model '/tmp/RtmpKix4lZ/file40e966e51ba7'
## Saved parameters: B Rho EnvRho
## MCMC ran in parallel for 2079.661 minutes at time 2019-04-10 08:26:21.
##
## For each of 5 chains:
## Adaptation: 250000 iterations (possibly insufficient)
## Burn-in: 0 iterations
## Thin rate: 100 iterations
## Total chain length: 270000 iterations
## Posterior sample size: 200 draws
##
## **WARNING** Rhat values indicate convergence failure.jsdm_conv(me20)## Maximum Rhat value for Beta: 1.2174
## Maximum Rhat value for Rho: NA
## Maximum Rhat value for EnvRho: 1.3236
me20$mcmc.info[1:7]## $n.chains
## [1] 5
##
## $n.adapt
## [1] 250000 250000 250000 250000 250000
##
## $sufficient.adapt
## [1] FALSE FALSE FALSE FALSE FALSE
##
## $n.iter
## [1] 20000
##
## $n.burnin
## [1] 0
##
## $n.thin
## [1] 100
##
## $n.samples
## [1] 1000data<-sim_data$EnvEvenSp20
data <- list(
Y = subset(data, select = -env),
X = cbind(1, scale(poly(data$env, 2))),
covx = cov(cbind(1, scale(poly(data$env, 2)))),
K = 3,
J = ncol(data) - 1,
n = nrow(data),
I = diag(ncol(data) - 1),
df = ncol(data)
)
##########################################################################################
plot_cor_jsdm(me20,data$Y)
Gjam
data<-sim_data$EnvEvenSp20
#fit_gjam(data,5000,500,"./gjam_models/gjam20env.rda")
load_gjam(data,name="./gjam_models/gjam20env.rda")##
## Sensitivity by predictor variables f:
## Estimate SE CI_025 CI_975
## env 335 61.7 222 461
## env2 157 28.3 110 219
##
## Coefficient matrix B:
## sp01 sp02 sp03 sp04 sp05 sp06 sp07 sp08 sp09
## intercept -0.533 0.0688 0.500 0.144 0.327 0.498 0.720 0.896 0.906
## env -1.850 -2.5200 -2.480 -2.020 -1.820 -1.760 -1.220 -0.753 -0.516
## env2 0.187 0.3400 0.451 -0.279 -0.507 -0.493 -0.651 -0.641 -0.815
## RMSPE 0.337 0.3000 0.309 0.331 0.332 0.333 0.365 0.398 0.406
## sp10 sp11 sp12 sp13 sp14 sp15 sp16 sp17 sp18
## intercept 1.130 1.120 1.050 0.689 0.610 0.435 0.440 0.113 -0.07100
## env -0.200 0.133 0.551 0.870 1.140 1.730 2.460 1.990 2.35000
## env2 -0.931 -0.855 -0.860 -0.794 -0.671 -0.452 -0.227 -0.271 0.00271
## RMSPE 0.398 0.411 0.386 0.399 0.363 0.339 0.308 0.330 0.30300
## sp19 sp20
## intercept -0.0697 -0.329
## env 2.5800 2.190
## env2 0.3120 0.530
## RMSPE 0.3030 0.318
##
## Coefficient matrix B:
## Estimate SE CI_025 CI_975 sig95
## sp01_intercept -0.53300 0.1340 -0.764000 -0.3250 *
## sp01_env -1.85000 0.1340 -2.100000 -1.5800 *
## sp01_env2 0.18700 0.0717 0.055600 0.3320 *
## sp02_intercept 0.06880 0.0972 -0.095900 0.2760
## sp02_env -2.52000 0.1560 -2.810000 -2.2200 *
## sp02_env2 0.34000 0.0720 0.206000 0.4840 *
## sp03_intercept 0.50000 0.1040 0.295000 0.6900 *
## sp03_env -2.48000 0.1360 -2.770000 -2.2400 *
## sp03_env2 0.45100 0.0813 0.310000 0.6170 *
## sp04_intercept 0.14400 0.0783 0.009940 0.3210 *
## sp04_env -2.02000 0.0913 -2.200000 -1.8400 *
## sp04_env2 -0.27900 0.0825 -0.432000 -0.1160 *
## sp05_intercept 0.32700 0.0843 0.172000 0.4850 *
## sp05_env -1.82000 0.1010 -2.020000 -1.6200 *
## sp05_env2 -0.50700 0.0716 -0.653000 -0.3720 *
## sp06_intercept 0.49800 0.0786 0.342000 0.6400 *
## sp06_env -1.76000 0.0786 -1.920000 -1.6100 *
## sp06_env2 -0.49300 0.0669 -0.631000 -0.3680 *
## sp07_intercept 0.72000 0.0628 0.598000 0.8430 *
## sp07_env -1.22000 0.0844 -1.370000 -1.0500 *
## sp07_env2 -0.65100 0.0816 -0.819000 -0.4980 *
## sp08_intercept 0.89600 0.0702 0.766000 1.0400 *
## sp08_env -0.75300 0.0678 -0.886000 -0.6230 *
## sp08_env2 -0.64100 0.0614 -0.759000 -0.5180 *
## sp09_intercept 0.90600 0.0712 0.776000 1.0500 *
## sp09_env -0.51600 0.0835 -0.663000 -0.3500 *
## sp09_env2 -0.81500 0.0669 -0.942000 -0.6790 *
## sp10_intercept 1.13000 0.0655 1.000000 1.2600 *
## sp10_env -0.20000 0.0573 -0.305000 -0.0862 *
## sp10_env2 -0.93100 0.0676 -1.060000 -0.8010 *
## sp11_intercept 1.12000 0.0684 0.987000 1.2500 *
## sp11_env 0.13300 0.0565 0.022800 0.2410 *
## sp11_env2 -0.85500 0.0630 -0.978000 -0.7310 *
## sp12_intercept 1.05000 0.0730 0.906000 1.2000 *
## sp12_env 0.55100 0.0747 0.420000 0.7010 *
## sp12_env2 -0.86000 0.0752 -1.010000 -0.7180 *
## sp13_intercept 0.68900 0.0591 0.570000 0.8000 *
## sp13_env 0.87000 0.0813 0.709000 1.0300 *
## sp13_env2 -0.79400 0.0668 -0.922000 -0.6560 *
## sp14_intercept 0.61000 0.0641 0.481000 0.7300 *
## sp14_env 1.14000 0.0675 1.020000 1.2800 *
## sp14_env2 -0.67100 0.0665 -0.800000 -0.5470 *
## sp15_intercept 0.43500 0.0588 0.325000 0.5530 *
## sp15_env 1.73000 0.1250 1.520000 1.9800 *
## sp15_env2 -0.45200 0.0641 -0.579000 -0.3340 *
## sp16_intercept 0.44000 0.0521 0.337000 0.5420 *
## sp16_env 2.46000 0.1050 2.260000 2.6700 *
## sp16_env2 -0.22700 0.0501 -0.323000 -0.1280 *
## sp17_intercept 0.11300 0.0563 0.000618 0.2230 *
## sp17_env 1.99000 0.1310 1.760000 2.2600 *
## sp17_env2 -0.27100 0.0645 -0.394000 -0.1480 *
## sp18_intercept -0.07100 0.0549 -0.181000 0.0357
## sp18_env 2.35000 0.1040 2.140000 2.5500 *
## sp18_env2 0.00271 0.0659 -0.130000 0.1260
## sp19_intercept -0.06970 0.0875 -0.241000 0.0747
## sp19_env 2.58000 0.1220 2.350000 2.8300 *
## sp19_env2 0.31200 0.0517 0.216000 0.4120 *
## sp20_intercept -0.32900 0.0556 -0.433000 -0.2160 *
## sp20_env 2.19000 0.0953 2.010000 2.3800 *
## sp20_env2 0.53000 0.0616 0.409000 0.6490 *
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Coefficient matrix B, standardized for X:
## NULL
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Coefficient matrix B, standardized for X and W:
## NULL
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Design Table
## env env2
## VIF 1 1
## env2 0 NA
##
## Sample contains n = 500 observations on S = 20 response variables, and 2 predictors. Data types (typeNames) include PA. There are 0 missing values in X and 0 missing values in Y. The RMSPE is 0.351, and the DIC is 371075. Computation involved 5000 Gibbs steps, with a burnin of 500.
#gje20<-load_object("./gjam_models/gjam20env.rda")
#to check posterior density of s in Sigma
#gje20<-load_object("./gjam_models/gjam20env.rda")
#plot(density(gje20$chains$sgibbs[,4]))Gjam dimension reduction
data<-sim_data$EnvEvenSp20
#fit_gjam(data,5000,500,"./gjam_models/gjam20env.rda")
#load_gjam(data,name="./gjam_models/gjam20env.rda")
#gje20<-load_object("./gjam_models/gjam20env.rda")
#to check posterior density of s in Sigma
#gje20<-load_object("./gjam_models/gjam20env.rda")
#plot(density(gje20$chains$sgibbs[,4]))
data <- list(
Y = subset(data, select = -env),
X = cbind(1, scale(poly(data$env, 2))),
covx = cov(cbind(1, scale(poly(data$env, 2)))),
K = 3,
J = ncol(data) - 1,
n = nrow(data),
I = diag(ncol(data) - 1),
df = ncol(data)
)
xdata<-as.data.frame(data$X[,-1])
colnames(xdata)<- c("env","env2")
ydata<-as.data.frame(data$Y)
###formula
rl <- list(r = 8, N = 20)
formula<-as.formula( ~env+ env2)
ml <- list(ng = 2500, burnin = 500, typeNames = 'PA', reductList = rl)
####fit
mod_gjam1 <- gjam(formula, xdata = xdata, ydata = ydata, modelList = ml)##
## Observations and responses:
## [1] 500 20## Warning in .setupReduct(modelList, S, Q, n): dimension reduction requires
## reductList$N < no. responses##
## Dimension reduced from 20 X 20 -> 20 X 8 responses
## ===========================================================================## expanding covariance chains## ===========================================================================save(mod_gjam1, file = "./gjam_models/gjam20env_dr.rda")
summary(mod_gjam1)##
## Sensitivity by predictor variables f:
## Estimate SE CI_025 CI_975
## env 1.12 0.0522 1.02 1.23
## env2 1.31 0.0608 1.19 1.43
##
## Coefficient matrix B:
## sp01 sp02 sp03 sp04 sp05 sp06 sp07 sp08
## intercept -0.331 -0.0900 0.023700 0.126 0.305 0.352 0.472 0.481
## env -0.476 -0.4830 -0.482000 -0.477 -0.472 -0.465 -0.375 -0.433
## env2 0.280 0.0941 -0.000526 -0.110 -0.264 -0.327 -0.436 -0.456
## RMSPE 0.405 0.4040 0.407000 0.410 0.402 0.402 0.412 0.419
## sp09 sp10 sp11 sp12 sp13 sp14 sp15 sp16 sp17
## intercept 0.485 0.489 0.489 0.487 0.475 0.478 0.378 0.252 0.153
## env -0.430 -0.106 0.321 0.463 0.448 0.401 0.469 0.479 0.480
## env2 -0.478 -0.484 -0.475 -0.481 -0.478 -0.455 -0.346 -0.256 -0.167
## RMSPE 0.429 0.441 0.449 0.428 0.423 0.410 0.406 0.406 0.408
## sp18 sp19 sp20
## intercept -0.0196 -0.117 -0.301
## env 0.4830 0.481 0.475
## env2 0.0111 0.104 0.246
## RMSPE 0.4030 0.406 0.403
##
## Coefficient matrix B:
## Estimate SE CI_025 CI_975 sig95
## sp01_intercept -0.331000 0.0699 -0.4640 -0.1830 *
## sp01_env -0.476000 0.0253 -0.4990 -0.4070 *
## sp01_env2 0.280000 0.0683 0.1400 0.4080 *
## sp02_intercept -0.090000 0.0703 -0.2290 0.0494
## sp02_env -0.483000 0.0170 -0.5000 -0.4400 *
## sp02_env2 0.094100 0.0702 -0.0424 0.2330
## sp03_intercept 0.023700 0.0720 -0.1160 0.1680
## sp03_env -0.482000 0.0188 -0.5000 -0.4280 *
## sp03_env2 -0.000526 0.0686 -0.1320 0.1310
## sp04_intercept 0.126000 0.0679 -0.0115 0.2540
## sp04_env -0.477000 0.0233 -0.4990 -0.4110 *
## sp04_env2 -0.110000 0.0690 -0.2490 0.0219
## sp05_intercept 0.305000 0.0714 0.1670 0.4470 *
## sp05_env -0.472000 0.0273 -0.4990 -0.4010 *
## sp05_env2 -0.264000 0.0721 -0.4070 -0.1240 *
## sp06_intercept 0.352000 0.0666 0.2200 0.4730 *
## sp06_env -0.465000 0.0328 -0.4990 -0.3790 *
## sp06_env2 -0.327000 0.0669 -0.4550 -0.1900 *
## sp07_intercept 0.472000 0.0266 0.4000 0.4990 *
## sp07_env -0.375000 0.0858 -0.4980 -0.2080 *
## sp07_env2 -0.436000 0.0471 -0.4970 -0.3270 *
## sp08_intercept 0.481000 0.0185 0.4290 0.4990 *
## sp08_env -0.433000 0.0536 -0.4980 -0.2990 *
## sp08_env2 -0.456000 0.0364 -0.4990 -0.3660 *
## sp09_intercept 0.485000 0.0151 0.4440 0.5000 *
## sp09_env -0.430000 0.0539 -0.4970 -0.3020 *
## sp09_env2 -0.478000 0.0206 -0.4990 -0.4230 *
## sp10_intercept 0.489000 0.0115 0.4560 0.5000 *
## sp10_env -0.106000 0.0861 -0.2490 0.1060
## sp10_env2 -0.484000 0.0157 -0.5000 -0.4420 *
## sp11_intercept 0.489000 0.0114 0.4600 0.5000 *
## sp11_env 0.321000 0.0707 0.1720 0.4540 *
## sp11_env2 -0.475000 0.0220 -0.4990 -0.4160 *
## sp12_intercept 0.487000 0.0129 0.4540 0.5000 *
## sp12_env 0.463000 0.0341 0.3710 0.4990 *
## sp12_env2 -0.481000 0.0175 -0.4990 -0.4350 *
## sp13_intercept 0.475000 0.0236 0.4100 0.4990 *
## sp13_env 0.448000 0.0495 0.3200 0.4990 *
## sp13_env2 -0.478000 0.0216 -0.5000 -0.4200 *
## sp14_intercept 0.478000 0.0208 0.4220 0.4990 *
## sp14_env 0.401000 0.0619 0.2730 0.4950 *
## sp14_env2 -0.455000 0.0383 -0.4990 -0.3570 *
## sp15_intercept 0.378000 0.0661 0.2360 0.4880 *
## sp15_env 0.469000 0.0279 0.3980 0.4990 *
## sp15_env2 -0.346000 0.0674 -0.4720 -0.2150 *
## sp16_intercept 0.252000 0.0670 0.1220 0.3810 *
## sp16_env 0.479000 0.0202 0.4230 0.4990 *
## sp16_env2 -0.256000 0.0679 -0.3850 -0.1240 *
## sp17_intercept 0.153000 0.0681 0.0191 0.2820 *
## sp17_env 0.480000 0.0205 0.4230 0.4990 *
## sp17_env2 -0.167000 0.0678 -0.2990 -0.0378 *
## sp18_intercept -0.019600 0.0713 -0.1580 0.1240
## sp18_env 0.483000 0.0172 0.4370 0.5000 *
## sp18_env2 0.011100 0.0713 -0.1280 0.1490
## sp19_intercept -0.117000 0.0701 -0.2560 0.0199
## sp19_env 0.481000 0.0187 0.4310 0.5000 *
## sp19_env2 0.104000 0.0683 -0.0272 0.2420
## sp20_intercept -0.301000 0.0693 -0.4320 -0.1640 *
## sp20_env 0.475000 0.0239 0.4120 0.4990 *
## sp20_env2 0.246000 0.0702 0.1100 0.3820 *
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Coefficient matrix B, standardized for X:
## NULL
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Coefficient matrix B, standardized for X and W:
## NULL
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Design Table
## env env2
## VIF 1 1
## env2 0 NA
##
## Sample contains n = 500 observations on S = 20 response variables, and 2 predictors. Data types (typeNames) include PA. There are 0 missing values in X and 0 missing values in Y. The RMSPE is 0.414, and the DIC is 77425. Computation involved 2500 Gibbs steps, with a burnin of 500. Dimension reduction was implemented with N = 9 and r = 8.Tau <- solve(mod_gjam1$parameters$sigMu)
Tau_n = to_prec(Tau)
#postH<-apply(mod_gjam1$chains$sgibbs, 2, quantile,0.95)
#postL<-apply(mod_gjam1$chains$sgibbs, 2, quantile,0.05)
#pH<-convert_to_m(postH)
#pL<-convert_to_m(postL)
#R_sign<-cov2cor(mod_gjam1$parameters$sigMu)*(!(pH>0 & pL<0))
par(mfrow=c(2,3),oma = c(1, 1, 1, 1))
corrplot(cor(ydata), diag = FALSE, order = "original",tl.pos = "ld", tl.cex = 0.5, method = "color",col=cols(200), type = "lower")
title("Correlation cor(Y)")
corrplot(mod_gjam1$parameters$corMu, diag = FALSE, order = "original",tl.pos = "ld", tl.cex = 0.5, method = "color",col=cols(200), type = "lower")
title("R")
corrplot(mod_gjam1$parameters$ematrix, diag = FALSE, order = "original",tl.pos = "ld", tl.cex = 0.5, method = "color",col=cols(200), type = "lower")
title("E matrix")
# corrplot(Tau_n, diag = FALSE, order = "original",tl.pos = "ld", tl.cex = 0.5, method = "color", type = "lower")
# title("Tau")
# corrplot(R_sign, diag = FALSE, order = "original",tl.pos = "ld", tl.cex = 0.5, method = "color", type = "lower")
# title("R signif")
corrplot(diag(20), diag = FALSE, order = "original",tl.pos = "ld", tl.cex = 0.5, method = "color",col=cols(200), type = "lower")
title("True interactions")
data<-sim_data$EnvEvenSp20
hm_mod<-fit_hmsc(data,"Load",name="./HMmodels/hm20env.rda" )
hm_conv(hm_mod)
hm_inter(hm_mod)
Environmental filtering + Facilitation dense 5 species
JSDM
mf5 <- load_object("model-2019-04-11-19-35-11.rda")
summary(mf5)## Summary for model '/tmp/RtmpKix4lZ/file40e94e482135'
## Saved parameters: B Rho EnvRho
## MCMC ran in parallel for 29.13 minutes at time 2019-04-11 19:06:03.
##
## For each of 5 chains:
## Adaptation: 250000 iterations (possibly insufficient)
## Burn-in: 0 iterations
## Thin rate: 100 iterations
## Total chain length: 270000 iterations
## Posterior sample size: 200 draws
##
## **WARNING** Rhat values indicate convergence failure.#mf5$Rhat
jsdm_conv(mf5)## Maximum Rhat value for Beta: 1.3081
## Maximum Rhat value for Rho: NA
## Maximum Rhat value for EnvRho: 1.1421
mf5$mcmc.info[1:7]## $n.chains
## [1] 5
##
## $n.adapt
## [1] 250000 250000 250000 250000 250000
##
## $sufficient.adapt
## [1] FALSE FALSE FALSE FALSE FALSE
##
## $n.iter
## [1] 20000
##
## $n.burnin
## [1] 0
##
## $n.thin
## [1] 100
##
## $n.samples
## [1] 1000data<-sim_data$FacDenseSp5
data <- list(
Y = subset(data, select = -env),
X = cbind(1, scale(poly(data$env, 2))),
covx = cov(cbind(1, scale(poly(data$env, 2)))),
K = 3,
J = ncol(data) - 1,
n = nrow(data),
I = diag(ncol(data) - 1),
df = ncol(data)
)
##########################################################################################
#Tau<-solve
#Tau_n<-to_prec(me10$mean$Tau)
#Tau_k<-Tau_n*(!(model$q97.5$Tau>0 & model$q2.5$Tau<0))
plot_cor_jsdm(mf5,data$Y,fac_inter[[4]])
Gjam
data<-sim_data$FacDenseSp5
#fit_gjam(data,10000,2000,"./gjam_models/gjam5f.rda",interact=fac_inter[[4]])
load_gjam(data,name="./gjam_models/gjam5f.rda", interact=fac_inter[[4]])##
## Sensitivity by predictor variables f:
## Estimate SE CI_025 CI_975
## env 27.6 13.50 8.15 58.8
## env2 18.4 6.44 8.83 34.3
##
## Coefficient matrix B:
## sp01 sp02 sp03 sp04 sp05
## intercept -1.640 0.979 3.170 1.890 -1.840
## env -1.750 -1.580 0.253 2.970 1.740
## env2 0.295 -0.575 -1.280 0.446 0.289
## RMSPE 0.310 0.312 0.204 0.326 0.308
##
## Coefficient matrix B:
## Estimate SE CI_025 CI_975 sig95
## sp01_intercept -1.640 0.1120 -1.910 -1.450 *
## sp01_env -1.750 0.1450 -2.040 -1.490 *
## sp01_env2 0.295 0.0642 0.171 0.418 *
## sp02_intercept 0.979 0.1040 0.774 1.190 *
## sp02_env -1.580 0.1390 -1.850 -1.330 *
## sp02_env2 -0.575 0.1090 -0.761 -0.347 *
## sp03_intercept 3.170 0.2530 2.730 3.680 *
## sp03_env 0.253 0.0773 0.110 0.402 *
## sp03_env2 -1.280 0.1260 -1.560 -1.050 *
## sp04_intercept 1.890 0.1980 1.450 2.230 *
## sp04_env 2.970 0.2390 2.500 3.450 *
## sp04_env2 0.446 0.1150 0.220 0.645 *
## sp05_intercept -1.840 0.1670 -2.210 -1.570 *
## sp05_env 1.740 0.1260 1.500 1.990 *
## sp05_env2 0.289 0.0699 0.141 0.425 *
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Coefficient matrix B, standardized for X:
## NULL
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Coefficient matrix B, standardized for X and W:
## NULL
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Design Table
## env env2
## VIF 1 1
## env2 0 NA
##
## Sample contains n = 500 observations on S = 5 response variables, and 2 predictors. Data types (typeNames) include PA. There are 0 missing values in X and 0 missing values in Y. The RMSPE is 0.296, and the DIC is 45158. Computation involved 10000 Gibbs steps, with a burnin of 2000.
#gjfd5<-load_object("./gjam_models/gjam5f.rda")
#to check posterior density of s in Sigma
#gjfd5<-load_object("./gjam_models/gjam5f.rda")
#plot(density(gjfd5$chains$sgibbs[,4]))HMSC
data<-sim_data$FacDenseSp5
hm_mod<-fit_hmsc(data,"Load",name="./HMmodels/hm5fd.rda" )
hm_conv(hm_mod)
hm_inter(hm_mod, interact = fac_inter[[4]])
Environmental filtering + Facilitation dense 10 species
JSDM
## Summary for model '/tmp/RtmpKix4lZ/file40e96d09e31e'
## Saved parameters: B Rho EnvRho
## MCMC ran in parallel for 684.487 minutes at time 2019-04-11 19:35:12.
##
## For each of 5 chains:
## Adaptation: 250000 iterations (possibly insufficient)
## Burn-in: 0 iterations
## Thin rate: 100 iterations
## Total chain length: 270000 iterations
## Posterior sample size: 200 draws
##
## **WARNING** Rhat values indicate convergence failure.## Maximum Rhat value for Beta: 1.2866
## Maximum Rhat value for Rho: NA
## Maximum Rhat value for EnvRho: 1.4808
## $n.chains
## [1] 5
##
## $n.adapt
## [1] 250000 250000 250000 250000 250000
##
## $sufficient.adapt
## [1] FALSE FALSE FALSE FALSE FALSE
##
## $n.iter
## [1] 20000
##
## $n.burnin
## [1] 0
##
## $n.thin
## [1] 100
##
## $n.samples
## [1] 1000
data<-sim_data$FacDenseSp10
#fit_gjam(data,5000,500,"./gjam_models/gjam10fd.rda",interact=fac_inter[[5]])
load_gjam(data,name="./gjam_models/gjam10fd.rda", interact=fac_inter[[5]])##
## Sensitivity by predictor variables f:
## Estimate SE CI_025 CI_975
## env 160.0 35.60 101.0 240
## env2 18.5 4.88 10.4 29
##
## Coefficient matrix B:
## sp01 sp02 sp03 sp04 sp05 sp06 sp07 sp08 sp09
## intercept 0.212 -1.490 -0.877 2.410 -0.0951 2.400 0.497 -0.401 1.260
## env -2.760 -2.660 -1.810 -3.010 -0.5910 1.160 1.330 1.030 3.990
## env2 1.070 -0.969 -0.644 0.319 -0.7020 -0.454 -0.256 -0.347 0.095
## RMSPE 0.274 0.375 0.406 0.305 0.4930 0.259 0.390 0.471 0.248
## sp10
## intercept -0.893
## env 2.250
## env2 -0.411
## RMSPE 0.354
##
## Coefficient matrix B:
## Estimate SE CI_025 CI_975 sig95
## sp01_intercept 0.2120 0.0687 0.0833 0.3520 *
## sp01_env -2.7600 0.1740 -3.1200 -2.4700 *
## sp01_env2 1.0700 0.1350 0.8190 1.3200 *
## sp02_intercept -1.4900 0.2000 -1.9200 -1.1600 *
## sp02_env -2.6600 0.2110 -3.0900 -2.3000 *
## sp02_env2 -0.9690 0.1130 -1.2400 -0.7780 *
## sp03_intercept -0.8770 0.1070 -1.0900 -0.6900 *
## sp03_env -1.8100 0.1470 -2.1200 -1.5700 *
## sp03_env2 -0.6440 0.1190 -0.9250 -0.4570 *
## sp04_intercept 2.4100 0.2040 2.0500 2.8000 *
## sp04_env -3.0100 0.2470 -3.4600 -2.5900 *
## sp04_env2 0.3190 0.0696 0.1860 0.4540 *
## sp05_intercept -0.0951 0.0554 -0.2050 0.0147
## sp05_env -0.5910 0.0628 -0.7140 -0.4700 *
## sp05_env2 -0.7020 0.0743 -0.8490 -0.5590 *
## sp06_intercept 2.4000 0.1260 2.1600 2.6400 *
## sp06_env 1.1600 0.0761 1.0100 1.3100 *
## sp06_env2 -0.4540 0.0649 -0.5840 -0.3310 *
## sp07_intercept 0.4970 0.0563 0.3840 0.6040 *
## sp07_env 1.3300 0.0797 1.1800 1.4900 *
## sp07_env2 -0.2560 0.0554 -0.3670 -0.1470 *
## sp08_intercept -0.4010 0.0539 -0.5070 -0.2950 *
## sp08_env 1.0300 0.0877 0.8540 1.1900 *
## sp08_env2 -0.3470 0.0592 -0.4640 -0.2320 *
## sp09_intercept 1.2600 0.0922 1.0900 1.4500 *
## sp09_env 3.9900 0.3810 3.3800 4.6800 *
## sp09_env2 0.0950 0.0613 -0.0361 0.2050
## sp10_intercept -0.8930 0.0796 -1.0700 -0.7550 *
## sp10_env 2.2500 0.1280 2.0400 2.5200 *
## sp10_env2 -0.4110 0.0576 -0.5230 -0.2980 *
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Coefficient matrix B, standardized for X:
## NULL
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Coefficient matrix B, standardized for X and W:
## NULL
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Design Table
## env env2
## VIF 1 1
## env2 0 NA
##
## Sample contains n = 500 observations on S = 10 response variables, and 2 predictors. Data types (typeNames) include PA. There are 0 missing values in X and 0 missing values in Y. The RMSPE is 0.367, and the DIC is 157914. Computation involved 5000 Gibbs steps, with a burnin of 500.
#gjfd5<-load_object("./gjam_models/gjam10fd.rda")
#to check posterior density of s in Sigma
#gjfd5<-load_object("./gjam_models/gjam10fd.rda")
#plot(density(gjfd5$chains$sgibbs[,4]))HMSC
Environmental filtering + Facilitation dense 20 species
JSDM
## Summary for model '/tmp/RtmpKix4lZ/file40e91ec9ff9'
## Saved parameters: B Rho EnvRho
## MCMC ran in parallel for 2062.467 minutes at time 2019-04-12 06:59:42.
##
## For each of 5 chains:
## Adaptation: 250000 iterations (possibly insufficient)
## Burn-in: 0 iterations
## Thin rate: 100 iterations
## Total chain length: 270000 iterations
## Posterior sample size: 200 draws
##
## **WARNING** Rhat values indicate convergence failure.## Maximum Rhat value for Beta: 1.6132
## Maximum Rhat value for Rho: NA
## Maximum Rhat value for EnvRho: 1.508
## $n.chains
## [1] 5
##
## $n.adapt
## [1] 250000 250000 250000 250000 250000
##
## $sufficient.adapt
## [1] FALSE FALSE FALSE FALSE FALSE
##
## $n.iter
## [1] 20000
##
## $n.burnin
## [1] 0
##
## $n.thin
## [1] 100
##
## $n.samples
## [1] 1000
Gjam
data<-sim_data$FacDenseSp20
#fit_gjam(data,10000,1500,"./gjam_models/gjam20fd.rda",interact=fac_inter[[6]])
load_gjam(data,name="./gjam_models/gjam20fd.rda", interact=fac_inter[[6]])##
## Sensitivity by predictor variables f:
## Estimate SE CI_025 CI_975
## env 301 55.7 207 421
## env2 170 30.7 121 243
##
## Coefficient matrix B:
## sp01 sp02 sp03 sp04 sp05 sp06 sp07 sp08 sp09
## intercept -0.528 -1.010 -0.866 0.194 -0.280 0.327 0.0883 0.310 0.494
## env -1.940 -1.950 -2.150 -2.820 -1.560 -2.170 -0.9060 -0.695 -0.778
## env2 0.294 -0.248 -0.681 -0.305 -0.784 -0.841 -0.6030 -0.903 -1.080
## RMSPE 0.323 0.362 0.371 0.293 0.401 0.285 0.4550 0.443 0.397
## sp10 sp11 sp12 sp13 sp14 sp15 sp16 sp17 sp18
## intercept 0.696 0.396 0.315 0.525 0.118 -0.218 -0.186 0.336 0.00648
## env -0.303 0.104 0.464 0.896 1.160 1.780 2.140 2.240 2.44000
## env2 -1.050 -0.883 -0.877 -0.915 -1.010 -1.260 -0.574 0.152 0.08790
## RMSPE 0.418 0.468 0.455 0.397 0.404 0.365 0.329 0.323 0.29800
## sp19 sp20
## intercept -0.0434 -0.393
## env 3.1700 1.980
## env2 0.5820 0.270
## RMSPE 0.2720 0.320
##
## Coefficient matrix B:
## Estimate SE CI_025 CI_975 sig95
## sp01_intercept -0.52800 0.1010 -0.71400 -0.3420 *
## sp01_env -1.94000 0.1660 -2.34000 -1.6800 *
## sp01_env2 0.29400 0.1980 0.02630 0.6830 *
## sp02_intercept -1.01000 0.1210 -1.25000 -0.7800 *
## sp02_env -1.95000 0.1570 -2.25000 -1.6600 *
## sp02_env2 -0.24800 0.1100 -0.44500 -0.0388 *
## sp03_intercept -0.86600 0.2110 -1.22000 -0.4330 *
## sp03_env -2.15000 0.1640 -2.44000 -1.7900 *
## sp03_env2 -0.68100 0.1130 -0.92200 -0.4750 *
## sp04_intercept 0.19400 0.0898 0.03490 0.3640 *
## sp04_env -2.82000 0.1770 -3.17000 -2.4900 *
## sp04_env2 -0.30500 0.0689 -0.43400 -0.1700 *
## sp05_intercept -0.28000 0.1190 -0.48000 -0.0615 *
## sp05_env -1.56000 0.1350 -1.80000 -1.3100 *
## sp05_env2 -0.78400 0.1100 -0.97000 -0.5580 *
## sp06_intercept 0.32700 0.0954 0.15300 0.5120 *
## sp06_env -2.17000 0.1190 -2.41000 -1.9500 *
## sp06_env2 -0.84100 0.1240 -1.15000 -0.6270 *
## sp07_intercept 0.08830 0.0641 -0.03460 0.2160
## sp07_env -0.90600 0.0843 -1.07000 -0.7410 *
## sp07_env2 -0.60300 0.0704 -0.74300 -0.4710 *
## sp08_intercept 0.31000 0.0662 0.18200 0.4390 *
## sp08_env -0.69500 0.0783 -0.84500 -0.5450 *
## sp08_env2 -0.90300 0.0732 -1.04000 -0.7580 *
## sp09_intercept 0.49400 0.0684 0.35900 0.6240 *
## sp09_env -0.77800 0.0880 -0.94700 -0.6100 *
## sp09_env2 -1.08000 0.0683 -1.21000 -0.9440 *
## sp10_intercept 0.69600 0.0602 0.57300 0.8090 *
## sp10_env -0.30300 0.0736 -0.44100 -0.1630 *
## sp10_env2 -1.05000 0.0715 -1.19000 -0.9050 *
## sp11_intercept 0.39600 0.0692 0.25800 0.5310 *
## sp11_env 0.10400 0.0612 -0.01520 0.2260
## sp11_env2 -0.88300 0.0692 -1.01000 -0.7440 *
## sp12_intercept 0.31500 0.0579 0.20300 0.4270 *
## sp12_env 0.46400 0.0548 0.35400 0.5690 *
## sp12_env2 -0.87700 0.0644 -0.99800 -0.7500 *
## sp13_intercept 0.52500 0.0640 0.40400 0.6530 *
## sp13_env 0.89600 0.0644 0.76900 1.0200 *
## sp13_env2 -0.91500 0.0909 -1.08000 -0.7390 *
## sp14_intercept 0.11800 0.0485 0.02210 0.2120 *
## sp14_env 1.16000 0.0696 1.02000 1.2900 *
## sp14_env2 -1.01000 0.0635 -1.13000 -0.8850 *
## sp15_intercept -0.21800 0.0620 -0.33500 -0.0928 *
## sp15_env 1.78000 0.1120 1.57000 1.9900 *
## sp15_env2 -1.26000 0.0795 -1.42000 -1.1100 *
## sp16_intercept -0.18600 0.1770 -0.47300 0.1040
## sp16_env 2.14000 0.1120 1.93000 2.3700 *
## sp16_env2 -0.57400 0.0868 -0.73500 -0.4050 *
## sp17_intercept 0.33600 0.0779 0.17800 0.4820 *
## sp17_env 2.24000 0.2570 1.87000 2.7400 *
## sp17_env2 0.15200 0.0798 0.00565 0.3220 *
## sp18_intercept 0.00648 0.0759 -0.13200 0.1440
## sp18_env 2.44000 0.1220 2.20000 2.6900 *
## sp18_env2 0.08790 0.0509 -0.01750 0.1850
## sp19_intercept -0.04340 0.0733 -0.16700 0.1070
## sp19_env 3.17000 0.1380 2.91000 3.4400 *
## sp19_env2 0.58200 0.0714 0.44700 0.7220 *
## sp20_intercept -0.39300 0.0627 -0.50600 -0.2600 *
## sp20_env 1.98000 0.1130 1.78000 2.2200 *
## sp20_env2 0.27000 0.0674 0.14400 0.4060 *
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Coefficient matrix B, standardized for X:
## NULL
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Coefficient matrix B, standardized for X and W:
## NULL
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Design Table
## env env2
## VIF 1 1
## env2 0 NA
##
## Sample contains n = 500 observations on S = 20 response variables, and 2 predictors. Data types (typeNames) include PA. There are 0 missing values in X and 0 missing values in Y. The RMSPE is 0.374, and the DIC is 400202. Computation involved 10000 Gibbs steps, with a burnin of 1500.
#gjfd5<-load_object("./gjam_models/gjam20fd.rda")
#to check posterior density of s in Sigma
#gjfd5<-load_object("./gjam_models/gjam20fd.rda")
#plot(density(gjfd5$chains$sgibbs[,4]))HMSC
data<-sim_data$FacDenseSp20
hm_mod<-fit_hmsc(data,"Load",name="./HMmodels/hm20fd.rda" )
hm_conv(hm_mod)
hm_inter(hm_mod, interact = fac_inter[[6]])
Environmental filtering + Facilitation sparse 5 species
JSDM
## Summary for model '/tmp/RtmpKix4lZ/file40e917d13dae'
## Saved parameters: B Rho EnvRho
## MCMC ran in parallel for 29.05 minutes at time 2019-04-13 17:22:13.
##
## For each of 5 chains:
## Adaptation: 250000 iterations (possibly insufficient)
## Burn-in: 0 iterations
## Thin rate: 100 iterations
## Total chain length: 270000 iterations
## Posterior sample size: 200 draws
##
## **WARNING** Rhat values indicate convergence failure.## Maximum Rhat value for Beta: 1.373
## Maximum Rhat value for Rho: NA
## Maximum Rhat value for EnvRho: 1.4048
## $n.chains
## [1] 5
##
## $n.adapt
## [1] 250000 250000 250000 250000 250000
##
## $sufficient.adapt
## [1] FALSE FALSE FALSE FALSE FALSE
##
## $n.iter
## [1] 20000
##
## $n.burnin
## [1] 0
##
## $n.thin
## [1] 100
##
## $n.samples
## [1] 1000
Gjam
data<-sim_data$FacSparseSp5
#fit_gjam(data,2500,500,"./gjam_models/gjam5fs.rda",interact=fac_inter[[7]])
load_gjam(data,name="./gjam_models/gjam5fs.rda", interact=fac_inter[[7]])##
## Sensitivity by predictor variables f:
## Estimate SE CI_025 CI_975
## env 14.0 8.42 4.23 35.1
## env2 7.8 3.77 3.27 18.2
##
## Coefficient matrix B:
## sp01 sp02 sp03 sp04 sp05
## intercept -0.209 0.212 0.308 1.030 0.0554
## env -2.850 -2.290 -0.671 2.060 4.3800
## env2 0.689 -0.895 -1.020 -0.513 -0.0713
## RMSPE 0.284 0.300 0.431 0.281 0.2490
##
## Coefficient matrix B:
## Estimate SE CI_025 CI_975 sig95
## sp01_intercept -0.2090 0.0925 -0.3770 -0.00693 *
## sp01_env -2.8500 0.2760 -3.4700 -2.45000 *
## sp01_env2 0.6890 0.1140 0.4510 0.90200 *
## sp02_intercept 0.2120 0.0582 0.0951 0.33100 *
## sp02_env -2.2900 0.1260 -2.5300 -2.05000 *
## sp02_env2 -0.8950 0.0888 -1.0600 -0.72600 *
## sp03_intercept 0.3080 0.0629 0.1830 0.43100 *
## sp03_env -0.6710 0.0731 -0.8110 -0.53600 *
## sp03_env2 -1.0200 0.0822 -1.1700 -0.85900 *
## sp04_intercept 1.0300 0.1100 0.8240 1.25000 *
## sp04_env 2.0600 0.1490 1.8200 2.39000 *
## sp04_env2 -0.5130 0.0616 -0.6320 -0.39400 *
## sp05_intercept 0.0554 0.1810 -0.2860 0.42500
## sp05_env 4.3800 0.6540 3.3400 5.72000 *
## sp05_env2 -0.0713 0.1730 -0.3960 0.28900
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Coefficient matrix B, standardized for X:
## NULL
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Coefficient matrix B, standardized for X and W:
## NULL
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Design Table
## env env2
## VIF 1 1
## env2 0 NA
##
## Sample contains n = 500 observations on S = 5 response variables, and 2 predictors. Data types (typeNames) include PA. There are 0 missing values in X and 0 missing values in Y. The RMSPE is 0.315, and the DIC is 18491. Computation involved 2500 Gibbs steps, with a burnin of 500.
#gjfd5<-load_object("./gjam_models/gjam5fs.rda")
#to check posterior density of s in Sigma
#gjfd5<-load_object("./gjam_models/gjam20fd.rda")
#plot(density(gjfd5$chains$sgibbs[,4]))HMSC
data<-sim_data$FacSparseSp5
hm_mod<-fit_hmsc(data,"Load",name="./HMmodels/hm5fs.rda" )
hm_conv(hm_mod)
hm_inter(hm_mod, interact = fac_inter[[7]])
Environmental filtering + Facilitation sparse 10 species
JSDM
## Summary for model '/tmp/RtmpKix4lZ/file40e9113ddc9'
## Saved parameters: B Rho EnvRho
## MCMC ran in parallel for 686.687 minutes at time 2019-04-13 17:51:16.
##
## For each of 5 chains:
## Adaptation: 250000 iterations (possibly insufficient)
## Burn-in: 0 iterations
## Thin rate: 100 iterations
## Total chain length: 270000 iterations
## Posterior sample size: 200 draws
##
## **WARNING** Rhat values indicate convergence failure.## Maximum Rhat value for Beta: 1.5479
## Maximum Rhat value for Rho: NA
## Maximum Rhat value for EnvRho: 1.4746
## $n.chains
## [1] 5
##
## $n.adapt
## [1] 250000 250000 250000 250000 250000
##
## $sufficient.adapt
## [1] FALSE FALSE FALSE FALSE FALSE
##
## $n.iter
## [1] 20000
##
## $n.burnin
## [1] 0
##
## $n.thin
## [1] 100
##
## $n.samples
## [1] 1000
data<-sim_data$FacSparseSp10
#fit_gjam(data,2500,500,"./gjam_models/gjam10fs.rda",interact=fac_inter[[8]])
load_gjam(data,name="./gjam_models/gjam10fs.rda", interact=fac_inter[[8]])##
## Sensitivity by predictor variables f:
## Estimate SE CI_025 CI_975
## env 162.0 39.3 92.4 247.0
## env2 59.7 12.6 39.3 88.8
##
## Coefficient matrix B:
## sp01 sp02 sp03 sp04 sp05 sp06 sp07 sp08 sp09
## intercept -1.140 0.492 0.732 0.052 0.278 0.268 0.108 0.9400 -0.888
## env -1.640 -2.630 -2.650 -0.992 -0.158 0.341 0.861 2.8900 1.840
## env2 -0.304 0.102 -0.413 -1.290 -1.130 -1.010 -0.984 -0.0437 -0.799
## RMSPE 0.412 0.297 0.267 0.408 0.434 0.451 0.436 0.2940 0.407
## sp10
## intercept -0.5640
## env 3.5400
## env2 -0.0422
## RMSPE 0.2560
##
## Coefficient matrix B:
## Estimate SE CI_025 CI_975 sig95
## sp01_intercept -1.1400 0.0911 -1.32000 -0.9660 *
## sp01_env -1.6400 0.1090 -1.83000 -1.4100 *
## sp01_env2 -0.3040 0.0872 -0.44700 -0.1160 *
## sp02_intercept 0.4920 0.0847 0.32600 0.6620 *
## sp02_env -2.6300 0.1210 -2.87000 -2.3900 *
## sp02_env2 0.1020 0.0618 -0.00800 0.2300
## sp03_intercept 0.7320 0.0855 0.57700 0.8950 *
## sp03_env -2.6500 0.1360 -2.94000 -2.4100 *
## sp03_env2 -0.4130 0.0611 -0.53900 -0.3000 *
## sp04_intercept 0.0520 0.0733 -0.08990 0.1830
## sp04_env -0.9920 0.0793 -1.17000 -0.8600 *
## sp04_env2 -1.2900 0.0841 -1.45000 -1.1200 *
## sp05_intercept 0.2780 0.0621 0.16600 0.4120 *
## sp05_env -0.1580 0.0607 -0.27700 -0.0445 *
## sp05_env2 -1.1300 0.0726 -1.28000 -0.9980 *
## sp06_intercept 0.2680 0.0596 0.15700 0.3860 *
## sp06_env 0.3410 0.0632 0.21300 0.4600 *
## sp06_env2 -1.0100 0.0733 -1.15000 -0.8710 *
## sp07_intercept 0.1080 0.0503 0.00799 0.2060 *
## sp07_env 0.8610 0.0894 0.70400 1.0300 *
## sp07_env2 -0.9840 0.0698 -1.12000 -0.8490 *
## sp08_intercept 0.9400 0.0703 0.79400 1.0700 *
## sp08_env 2.8900 0.1680 2.54000 3.2100 *
## sp08_env2 -0.0437 0.0504 -0.13900 0.0564
## sp09_intercept -0.8880 0.0713 -1.02000 -0.7410 *
## sp09_env 1.8400 0.0811 1.67000 1.9900 *
## sp09_env2 -0.7990 0.0645 -0.91600 -0.6650 *
## sp10_intercept -0.5640 0.0605 -0.67500 -0.4400 *
## sp10_env 3.5400 0.1800 3.23000 3.9300 *
## sp10_env2 -0.0422 0.0522 -0.14100 0.0606
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Coefficient matrix B, standardized for X:
## NULL
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Coefficient matrix B, standardized for X and W:
## NULL
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Design Table
## env env2
## VIF 1 1
## env2 0 NA
##
## Sample contains n = 500 observations on S = 10 response variables, and 2 predictors. Data types (typeNames) include PA. There are 0 missing values in X and 0 missing values in Y. The RMSPE is 0.374, and the DIC is 120194. Computation involved 2500 Gibbs steps, with a burnin of 500.
#gjfd5<-load_object("./gjam_models/gjam10fs.rda")
#to check posterior density of s in Sigma
#gjfd5<-load_object("./gjam_models/gjam10fs.rda")
#plot(density(gjfd5$chains$sgibbs[,4]))HMSC
data<-sim_data$FacSparseSp10
hm_mod<-fit_hmsc(data,"Load",name="./HMmodels/hm10fs.rda" )
hm_conv(hm_mod)
hm_inter(hm_mod, interact = fac_inter[[8]])
Environmental filtering + Facilitation sparse 20 species
JSDM
## Summary for model '/tmp/RtmpKix4lZ/file40e9629c97ad'
## Saved parameters: B Rho EnvRho
## MCMC ran in parallel for 2056.316 minutes at time 2019-04-14 05:17:59.
##
## For each of 5 chains:
## Adaptation: 250000 iterations (possibly insufficient)
## Burn-in: 0 iterations
## Thin rate: 100 iterations
## Total chain length: 270000 iterations
## Posterior sample size: 200 draws
##
## **WARNING** Rhat values indicate convergence failure.## Maximum Rhat value for Beta: 1.5653
## Maximum Rhat value for Rho: NA
## Maximum Rhat value for EnvRho: 1.4826
## $n.chains
## [1] 5
##
## $n.adapt
## [1] 250000 250000 250000 250000 250000
##
## $sufficient.adapt
## [1] FALSE FALSE FALSE FALSE FALSE
##
## $n.iter
## [1] 20000
##
## $n.burnin
## [1] 0
##
## $n.thin
## [1] 100
##
## $n.samples
## [1] 1000
Gjam
data<-sim_data$FacSparseSp20
#fit_gjam(data,10000,1500,"./gjam_models/gjam20fs.rda",interact=fac_inter[[9]])
load_gjam(data,name="./gjam_models/gjam20fs.rda", interact=fac_inter[[9]])##
## Sensitivity by predictor variables f:
## Estimate SE CI_025 CI_975
## env 300 65.9 188.0 443
## env2 148 31.5 94.4 218
##
## Coefficient matrix B:
## sp01 sp02 sp03 sp04 sp05 sp06 sp07 sp08 sp09
## intercept -0.330 -1.030 0.197 -0.0566 0.6180 0.691 0.485 0.555 0.880
## env -2.590 -2.240 -2.560 -2.1000 -1.7500 -1.690 -1.130 -0.915 -0.641
## env2 0.429 -0.362 0.131 -0.5870 -0.0122 -0.439 -0.559 -0.728 -0.860
## RMSPE 0.295 0.348 0.300 0.3050 0.3520 0.333 0.395 0.410 0.403
## sp10 sp11 sp12 sp13 sp14 sp15 sp16 sp17 sp18
## intercept 0.862 0.847 0.601 0.700 0.447 0.594 0.337 0.302 0.425
## env -0.212 0.236 0.513 1.050 0.987 1.530 1.520 2.750 2.190
## env2 -0.988 -1.140 -0.824 -0.958 -0.741 -0.448 -0.281 -0.307 0.525
## RMSPE 0.410 0.401 0.438 0.363 0.408 0.341 0.364 0.277 0.320
## sp19 sp20
## intercept -0.105 -0.724
## env 2.440 2.280
## env2 0.281 -0.001
## RMSPE 0.292 0.304
##
## Coefficient matrix B:
## Estimate SE CI_025 CI_975 sig95
## sp01_intercept -0.3300 0.0646 -0.4540 -0.1970 *
## sp01_env -2.5900 0.1480 -2.8600 -2.2700 *
## sp01_env2 0.4290 0.1770 0.0651 0.6740 *
## sp02_intercept -1.0300 0.1010 -1.2200 -0.8380 *
## sp02_env -2.2400 0.1410 -2.5600 -2.0100 *
## sp02_env2 -0.3620 0.1120 -0.6120 -0.1510 *
## sp03_intercept 0.1970 0.0816 0.0411 0.3590 *
## sp03_env -2.5600 0.1810 -2.9200 -2.2200 *
## sp03_env2 0.1310 0.1050 -0.0655 0.3510
## sp04_intercept -0.0566 0.0891 -0.2220 0.1230
## sp04_env -2.1000 0.1340 -2.4000 -1.8700 *
## sp04_env2 -0.5870 0.0776 -0.7300 -0.4350 *
## sp05_intercept 0.6180 0.0920 0.4430 0.7790 *
## sp05_env -1.7500 0.1070 -1.9600 -1.5400 *
## sp05_env2 -0.0122 0.1040 -0.2690 0.1470
## sp06_intercept 0.6910 0.0806 0.5250 0.8370 *
## sp06_env -1.6900 0.1130 -1.9200 -1.4900 *
## sp06_env2 -0.4390 0.1030 -0.6550 -0.2760 *
## sp07_intercept 0.4850 0.0750 0.3440 0.6410 *
## sp07_env -1.1300 0.1010 -1.3300 -0.9460 *
## sp07_env2 -0.5590 0.1180 -0.7940 -0.3560 *
## sp08_intercept 0.5550 0.0615 0.4290 0.6730 *
## sp08_env -0.9150 0.0719 -1.0600 -0.7820 *
## sp08_env2 -0.7280 0.0797 -0.8670 -0.5640 *
## sp09_intercept 0.8800 0.0866 0.6900 1.0300 *
## sp09_env -0.6410 0.0609 -0.7610 -0.5230 *
## sp09_env2 -0.8600 0.1060 -1.0700 -0.6590 *
## sp10_intercept 0.8620 0.0572 0.7490 0.9730 *
## sp10_env -0.2120 0.0564 -0.3190 -0.0966 *
## sp10_env2 -0.9880 0.0745 -1.1400 -0.8510 *
## sp11_intercept 0.8470 0.0636 0.7210 0.9710 *
## sp11_env 0.2360 0.0671 0.1140 0.3790 *
## sp11_env2 -1.1400 0.0851 -1.2900 -0.9680 *
## sp12_intercept 0.6010 0.0696 0.4510 0.7280 *
## sp12_env 0.5130 0.0584 0.3980 0.6270 *
## sp12_env2 -0.8240 0.0719 -0.9680 -0.6890 *
## sp13_intercept 0.7000 0.0739 0.5550 0.8400 *
## sp13_env 1.0500 0.0776 0.9050 1.2000 *
## sp13_env2 -0.9580 0.0740 -1.1000 -0.8100 *
## sp14_intercept 0.4470 0.0761 0.2960 0.5960 *
## sp14_env 0.9870 0.0819 0.8290 1.1500 *
## sp14_env2 -0.7410 0.0739 -0.8830 -0.5950 *
## sp15_intercept 0.5940 0.0578 0.4820 0.7060 *
## sp15_env 1.5300 0.0960 1.3500 1.7100 *
## sp15_env2 -0.4480 0.0678 -0.5700 -0.3110 *
## sp16_intercept 0.3370 0.0589 0.2160 0.4480 *
## sp16_env 1.5200 0.0753 1.3700 1.6700 *
## sp16_env2 -0.2810 0.0615 -0.4050 -0.1670 *
## sp17_intercept 0.3020 0.1020 0.1030 0.4720 *
## sp17_env 2.7500 0.1330 2.4700 2.9900 *
## sp17_env2 -0.3070 0.0872 -0.4540 -0.1390 *
## sp18_intercept 0.4250 0.0906 0.2660 0.6020 *
## sp18_env 2.1900 0.1370 1.9400 2.4600 *
## sp18_env2 0.5250 0.0868 0.3550 0.6790 *
## sp19_intercept -0.1050 0.0792 -0.2810 0.0245
## sp19_env 2.4400 0.1240 2.2200 2.7000 *
## sp19_env2 0.2810 0.0574 0.1650 0.3890 *
## sp20_intercept -0.7240 0.0791 -0.8750 -0.5750 *
## sp20_env 2.2800 0.1140 2.0600 2.5000 *
## sp20_env2 -0.0010 0.0740 -0.1410 0.1360
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Coefficient matrix B, standardized for X:
## NULL
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Coefficient matrix B, standardized for X and W:
## NULL
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Design Table
## env env2
## VIF 1 1
## env2 0 NA
##
## Sample contains n = 500 observations on S = 20 response variables, and 2 predictors. Data types (typeNames) include PA. There are 0 missing values in X and 0 missing values in Y. The RMSPE is 0.356, and the DIC is 450253. Computation involved 10000 Gibbs steps, with a burnin of 1500.
#gjfd5<-load_object("./gjam_models/gjam20fd.rda")
#to check posterior density of s in Sigma
#gjfd5<-load_object("./gjam_models/gjam20fd.rda")
#plot(density(gjfd5$chains$sgibbs[,4]))HMSC
HMSC
data<-sim_data$FacSparseSp20
hm_mod<-fit_hmsc(data,nsamples=3000, nchains=2,"Load",name="./HMmodels/hm20fs.rda" )
hm_conv(hm_mod)
hm_inter(hm_mod, nsamples=3000, nchains=2,interact = fac_inter[[9]])
Environmental filtering + Competition dense 5 species
JSDM
## Summary for model '/tmp/RtmpKix4lZ/file40e9127bbb96'
## Saved parameters: B Rho EnvRho
## MCMC ran in parallel for 29.308 minutes at time 2019-04-15 15:34:20.
##
## For each of 5 chains:
## Adaptation: 250000 iterations (possibly insufficient)
## Burn-in: 0 iterations
## Thin rate: 100 iterations
## Total chain length: 270000 iterations
## Posterior sample size: 200 draws
##
## Successful convergence based on Rhat values (all < 1.1).## Maximum Rhat value for Beta: 1.0324
## Maximum Rhat value for Rho: NA
## Maximum Rhat value for EnvRho: 1.0313
## $n.chains
## [1] 5
##
## $n.adapt
## [1] 250000 250000 250000 250000 250000
##
## $sufficient.adapt
## [1] FALSE FALSE FALSE FALSE FALSE
##
## $n.iter
## [1] 20000
##
## $n.burnin
## [1] 0
##
## $n.thin
## [1] 100
##
## $n.samples
## [1] 1000
Gjam
data<-sim_data$CompDenseSp5
#fit_gjam(data,10000,1000,"./gjam_models/gjam5cmpd.rda",interact=comp_inter[[10]])
load_gjam(data,name="./gjam_models/gjam5cmpd.rda", interact=(-1)*comp_inter[[10]])##
## Sensitivity by predictor variables f:
## Estimate SE CI_025 CI_975
## env 73.8 28.4 22.3 132.0
## env2 43.7 19.0 17.4 89.3
##
## Coefficient matrix B:
## sp01 sp02 sp03 sp04 sp05
## intercept -1.340 0.170 1.7800 0.236 -1.420
## env -0.661 -0.863 0.0249 0.662 0.680
## env2 0.280 -1.080 -0.7440 -0.949 0.287
## RMSPE 0.351 0.427 0.3080 0.456 0.343
##
## Coefficient matrix B:
## Estimate SE CI_025 CI_975 sig95
## sp01_intercept -1.3400 0.0888 -1.5100 -1.170 *
## sp01_env -0.6610 0.0777 -0.8020 -0.500 *
## sp01_env2 0.2800 0.0782 0.1170 0.417 *
## sp02_intercept 0.1700 0.0748 0.0112 0.302 *
## sp02_env -0.8630 0.0689 -0.9970 -0.731 *
## sp02_env2 -1.0800 0.0847 -1.2500 -0.918 *
## sp03_intercept 1.7800 0.1300 1.5300 2.020 *
## sp03_env 0.0249 0.0622 -0.0989 0.147
## sp03_env2 -0.7440 0.1000 -0.9310 -0.551 *
## sp04_intercept 0.2360 0.0543 0.1300 0.342 *
## sp04_env 0.6620 0.0765 0.5340 0.851 *
## sp04_env2 -0.9490 0.0790 -1.1200 -0.807 *
## sp05_intercept -1.4200 0.0901 -1.6000 -1.240 *
## sp05_env 0.6800 0.0742 0.5440 0.831 *
## sp05_env2 0.2870 0.0746 0.1490 0.425 *
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Coefficient matrix B, standardized for X:
## NULL
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Coefficient matrix B, standardized for X and W:
## NULL
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Design Table
## env env2
## VIF 1 1
## env2 0 NA
##
## Sample contains n = 500 observations on S = 5 response variables, and 2 predictors. Data types (typeNames) include PA. There are 0 missing values in X and 0 missing values in Y. The RMSPE is 0.381, and the DIC is 31214. Computation involved 10000 Gibbs steps, with a burnin of 1000.
#gjfd5<-load_object("./gjam_models/gjam5cmpd.rda")
#to check posterior density of s in Sigma
#gjfd5<-load_object("./gjam_models/gjam5cmpd.rda")
#plot(density(gjfd5$chains$sgibbs[,4]))data<-sim_data$CompDenseSp5
hm_mod<-fit_hmsc(data,"Load",nsamples=3000, nchains=2,name="./HMmodels/hm5cmpd.rda" )
hm_conv(hm_mod)
hm_inter(hm_mod, nsamples=3000, nchains=2,interact = (-1)*comp_inter[[10]])
Environmental filtering + Competition dense 10 species
## Summary for model '/tmp/RtmpKix4lZ/file40e9559ce9d8'
## Saved parameters: B Rho EnvRho
## MCMC ran in parallel for 695.614 minutes at time 2019-04-15 16:03:39.
##
## For each of 5 chains:
## Adaptation: 250000 iterations (possibly insufficient)
## Burn-in: 0 iterations
## Thin rate: 100 iterations
## Total chain length: 270000 iterations
## Posterior sample size: 200 draws
##
## **WARNING** Rhat values indicate convergence failure.## Maximum Rhat value for Beta: 1.1845
## Maximum Rhat value for Rho: NA
## Maximum Rhat value for EnvRho: 1.1771
## $n.chains
## [1] 5
##
## $n.adapt
## [1] 250000 250000 250000 250000 250000
##
## $sufficient.adapt
## [1] FALSE FALSE FALSE FALSE FALSE
##
## $n.iter
## [1] 20000
##
## $n.burnin
## [1] 0
##
## $n.thin
## [1] 100
##
## $n.samples
## [1] 1000
Gjam
data<-sim_data$CompDenseSp10
#fit_gjam(data,5000,500,"./gjam_models/gjam10cmpd.rda",interact=comp_inter[[11]])
load_gjam(data,name="./gjam_models/gjam10cmpd.rda", interact= (-1)*comp_inter[[11]])##
## Sensitivity by predictor variables f:
## Estimate SE CI_025 CI_975
## env 88.6 22.30 52.1 138.0
## env2 40.9 9.14 24.7 60.8
##
## Coefficient matrix B:
## sp01 sp02 sp03 sp04 sp05 sp06 sp07 sp08 sp09
## intercept -0.123 -1.360 0.287 1.150 0.552 1.230 -0.933 0.459 0.629
## env -2.520 -0.489 -1.150 -1.210 -0.255 0.376 0.868 1.790 2.790
## env2 0.141 0.308 -0.995 -0.557 -0.577 -0.833 -0.830 -0.579 0.276
## RMSPE 0.297 0.341 0.389 0.327 0.493 0.385 0.450 0.318 0.301
## sp10
## intercept -0.519
## env 1.370
## env2 0.513
## RMSPE 0.359
##
## Coefficient matrix B:
## Estimate SE CI_025 CI_975 sig95
## sp01_intercept -0.123 0.0954 -0.2940 0.0708
## sp01_env -2.520 0.1750 -2.8800 -2.2100 *
## sp01_env2 0.141 0.1240 -0.0623 0.3830
## sp02_intercept -1.360 0.0758 -1.5100 -1.2100 *
## sp02_env -0.489 0.0590 -0.6020 -0.3750 *
## sp02_env2 0.308 0.0741 0.1760 0.4580 *
## sp03_intercept 0.287 0.0798 0.1340 0.4330 *
## sp03_env -1.150 0.0816 -1.3100 -1.0000 *
## sp03_env2 -0.995 0.0941 -1.1600 -0.8190 *
## sp04_intercept 1.150 0.0925 0.9740 1.3300 *
## sp04_env -1.210 0.1000 -1.4000 -1.0200 *
## sp04_env2 -0.557 0.0767 -0.7390 -0.4220 *
## sp05_intercept 0.552 0.0611 0.4330 0.6720 *
## sp05_env -0.255 0.0535 -0.3610 -0.1530 *
## sp05_env2 -0.577 0.0637 -0.7010 -0.4560 *
## sp06_intercept 1.230 0.0730 1.0900 1.3800 *
## sp06_env 0.376 0.0650 0.2530 0.4990 *
## sp06_env2 -0.833 0.0821 -0.9990 -0.6810 *
## sp07_intercept -0.933 0.0741 -1.0800 -0.7860 *
## sp07_env 0.868 0.0745 0.7100 1.0000 *
## sp07_env2 -0.830 0.0831 -1.0100 -0.6750 *
## sp08_intercept 0.459 0.0739 0.3310 0.6170 *
## sp08_env 1.790 0.1070 1.5900 2.0100 *
## sp08_env2 -0.579 0.0596 -0.6980 -0.4630 *
## sp09_intercept 0.629 0.0599 0.5130 0.7470 *
## sp09_env 2.790 0.1360 2.5400 3.0700 *
## sp09_env2 0.276 0.0597 0.1570 0.3880 *
## sp10_intercept -0.519 0.0622 -0.6470 -0.4040 *
## sp10_env 1.370 0.1000 1.1800 1.5600 *
## sp10_env2 0.513 0.0589 0.3890 0.6240 *
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Coefficient matrix B, standardized for X:
## NULL
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Coefficient matrix B, standardized for X and W:
## NULL
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Design Table
## env env2
## VIF 1 1
## env2 0 NA
##
## Sample contains n = 500 observations on S = 10 response variables, and 2 predictors. Data types (typeNames) include PA. There are 0 missing values in X and 0 missing values in Y. The RMSPE is 0.371, and the DIC is 83249. Computation involved 5000 Gibbs steps, with a burnin of 500.
#gjfd5<-load_object("./gjam_models/gjam10cmpd.rda")
#to check posterior density of s in Sigma
#gjfd5<-load_object("./gjam_models/gjam20fd.rda")
#plot(density(gjfd5$chains$sgibbs[,4]))HMSC
data<-sim_data$CompDenseSp10
hm_mod<-fit_hmsc(data,"Load",nsamples=5000, nchains=2,name="./HMmodels/hm10cmpd.rda" )
hm_conv(hm_mod)
hm_inter(hm_mod,nsamples=5000, nchains=2, interact = (-1)*comp_inter[[11]])
Environmental filtering + Competition dense 20 species
## Summary for model '/tmp/RtmpKix4lZ/file40e96843124a'
## Saved parameters: B Rho EnvRho
## MCMC ran in parallel for 2061.424 minutes at time 2019-04-16 03:39:17.
##
## For each of 5 chains:
## Adaptation: 250000 iterations (possibly insufficient)
## Burn-in: 0 iterations
## Thin rate: 100 iterations
## Total chain length: 270000 iterations
## Posterior sample size: 200 draws
##
## **WARNING** Rhat values indicate convergence failure.## Maximum Rhat value for Beta: 1.9065
## Maximum Rhat value for Rho: NA
## Maximum Rhat value for EnvRho: 1.6944
## $n.chains
## [1] 5
##
## $n.adapt
## [1] 250000 250000 250000 250000 250000
##
## $sufficient.adapt
## [1] FALSE FALSE FALSE FALSE FALSE
##
## $n.iter
## [1] 20000
##
## $n.burnin
## [1] 0
##
## $n.thin
## [1] 100
##
## $n.samples
## [1] 1000
Gjam
data<-sim_data$CompDenseSp20
#fit_gjam(data,10000,1000,"./gjam_models/gjam20cmpd.rda",interact= (-1)*comp_inter[[12]])
load_gjam(data,name="./gjam_models/gjam20cmpd.rda", interact= (-1)*comp_inter[[12]])##
## Sensitivity by predictor variables f:
## Estimate SE CI_025 CI_975
## env 318 54.6 218.0 432
## env2 116 20.8 78.2 160
##
## Coefficient matrix B:
## sp01 sp02 sp03 sp04 sp05 sp06 sp07 sp08 sp09
## intercept -0.0358 -1.130 -0.096 0.511 -0.184 0.893 0.0506 1.420 1.120
## env -2.1300 -1.070 -1.910 -2.900 -1.300 -1.630 -0.5920 -0.832 -0.395
## env2 0.4330 -0.470 -0.696 -0.240 -0.649 -0.591 -0.3470 -0.637 -0.657
## RMSPE 0.3040 0.453 0.348 0.291 0.438 0.316 0.5230 0.343 0.394
## sp10 sp11 sp12 sp13 sp14 sp15 sp16 sp17 sp18
## intercept 0.6690 0.0535 1.320 0.851 0.300 0.884 1.1300 -0.200 0.501
## env -0.0441 -0.0424 0.831 0.931 0.686 1.690 2.5100 1.460 2.410
## env2 -0.4730 -0.2160 -0.519 -0.597 -0.967 -0.316 0.0994 -0.107 0.467
## RMSPE 0.4950 0.5700 0.355 0.388 0.445 0.331 0.3100 0.411 0.326
## sp19 sp20
## intercept -0.4190 -2.040
## env 2.2000 0.776
## env2 -0.0511 -1.040
## RMSPE 0.3290 0.323
##
## Coefficient matrix B:
## Estimate SE CI_025 CI_975 sig95
## sp01_intercept -0.0358 0.0820 -0.19400 0.1180
## sp01_env -2.1300 0.1020 -2.33000 -1.9400 *
## sp01_env2 0.4330 0.0873 0.26500 0.5990 *
## sp02_intercept -1.1300 0.0847 -1.29000 -0.9620 *
## sp02_env -1.0700 0.0834 -1.24000 -0.9120 *
## sp02_env2 -0.4700 0.1200 -0.69000 -0.2480 *
## sp03_intercept -0.0960 0.0872 -0.26300 0.0815
## sp03_env -1.9100 0.1060 -2.13000 -1.7100 *
## sp03_env2 -0.6960 0.1170 -0.89800 -0.4690 *
## sp04_intercept 0.5110 0.0936 0.32900 0.6790 *
## sp04_env -2.9000 0.1400 -3.18000 -2.6300 *
## sp04_env2 -0.2400 0.2230 -0.62100 0.0949
## sp05_intercept -0.1840 0.0748 -0.33600 -0.0353 *
## sp05_env -1.3000 0.0866 -1.49000 -1.1400 *
## sp05_env2 -0.6490 0.1120 -0.89000 -0.4670 *
## sp06_intercept 0.8930 0.0924 0.70000 1.0600 *
## sp06_env -1.6300 0.1040 -1.84000 -1.4400 *
## sp06_env2 -0.5910 0.0673 -0.73300 -0.4680 *
## sp07_intercept 0.0506 0.0585 -0.06590 0.1630
## sp07_env -0.5920 0.0617 -0.71400 -0.4700 *
## sp07_env2 -0.3470 0.0604 -0.46100 -0.2210 *
## sp08_intercept 1.4200 0.0957 1.23000 1.5900 *
## sp08_env -0.8320 0.0847 -0.99900 -0.6660 *
## sp08_env2 -0.6370 0.0756 -0.78600 -0.4930 *
## sp09_intercept 1.1200 0.0746 0.97500 1.2700 *
## sp09_env -0.3950 0.0564 -0.50700 -0.2870 *
## sp09_env2 -0.6570 0.0945 -0.84500 -0.4880 *
## sp10_intercept 0.6690 0.0746 0.51100 0.8000 *
## sp10_env -0.0441 0.0569 -0.15200 0.0702
## sp10_env2 -0.4730 0.0697 -0.61300 -0.3430 *
## sp11_intercept 0.0535 0.0655 -0.07700 0.1790
## sp11_env -0.0424 0.0578 -0.15800 0.0683
## sp11_env2 -0.2160 0.0679 -0.34500 -0.0781 *
## sp12_intercept 1.3200 0.0891 1.14000 1.4800 *
## sp12_env 0.8310 0.0726 0.69500 0.9750 *
## sp12_env2 -0.5190 0.0947 -0.68700 -0.3440 *
## sp13_intercept 0.8510 0.0630 0.72600 0.9720 *
## sp13_env 0.9310 0.0732 0.77700 1.0700 *
## sp13_env2 -0.5970 0.0705 -0.73700 -0.4640 *
## sp14_intercept 0.3000 0.0604 0.19300 0.4350 *
## sp14_env 0.6860 0.0675 0.55500 0.8180 *
## sp14_env2 -0.9670 0.0751 -1.11000 -0.8230 *
## sp15_intercept 0.8840 0.0657 0.75500 1.0100 *
## sp15_env 1.6900 0.1250 1.48000 1.9600 *
## sp15_env2 -0.3160 0.0772 -0.46100 -0.1650 *
## sp16_intercept 1.1300 0.0883 0.97200 1.3200 *
## sp16_env 2.5100 0.1230 2.29000 2.7700 *
## sp16_env2 0.0994 0.0559 -0.00627 0.2120
## sp17_intercept -0.2000 0.0714 -0.33200 -0.0553 *
## sp17_env 1.4600 0.0885 1.29000 1.6300 *
## sp17_env2 -0.1070 0.0772 -0.26400 0.0315
## sp18_intercept 0.5010 0.0802 0.35400 0.6640 *
## sp18_env 2.4100 0.1350 2.14000 2.6600 *
## sp18_env2 0.4670 0.0959 0.29100 0.6590 *
## sp19_intercept -0.4190 0.0794 -0.57000 -0.2740 *
## sp19_env 2.2000 0.1060 2.00000 2.4100 *
## sp19_env2 -0.0511 0.0801 -0.20500 0.1060
## sp20_intercept -2.0400 0.1240 -2.28000 -1.8100 *
## sp20_env 0.7760 0.0763 0.63800 0.9350 *
## sp20_env2 -1.0400 0.0822 -1.20000 -0.8790 *
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Coefficient matrix B, standardized for X:
## NULL
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Coefficient matrix B, standardized for X and W:
## NULL
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Design Table
## env env2
## VIF 1 1
## env2 0 NA
##
## Sample contains n = 500 observations on S = 20 response variables, and 2 predictors. Data types (typeNames) include PA. There are 0 missing values in X and 0 missing values in Y. The RMSPE is 0.392, and the DIC is 431994. Computation involved 10000 Gibbs steps, with a burnin of 1000.
#gjfd5<-load_object("./gjam_models/gjam20cmpd.rda")
#to check posterior density of s in Sigma
#gjfd5<-load_object("./gjam_models/gjam20fd.rda")
#plot(density(gjfd5$chains$sgibbs[,4]))HMSC
data<-sim_data$CompDenseSp20
hm_mod<-fit_hmsc(data,"Load",nsamples=5000, nchains=2,name="./HMmodels/hm20cmpd.rda" )
hm_conv(hm_mod)
hm_inter(hm_mod, nsamples=5000, nchains=2,interact = (-1)*comp_inter[[12]])
Environmental filtering + Competition sparse 5 species
## Summary for model '/tmp/RtmpKix4lZ/file40e97e225117'
## Saved parameters: B Rho EnvRho
## MCMC ran in parallel for 29.251 minutes at time 2019-04-17 14:00:46.
##
## For each of 5 chains:
## Adaptation: 250000 iterations (possibly insufficient)
## Burn-in: 0 iterations
## Thin rate: 100 iterations
## Total chain length: 270000 iterations
## Posterior sample size: 200 draws
##
## **WARNING** Rhat values indicate convergence failure.## Maximum Rhat value for Beta: 1.1876
## Maximum Rhat value for Rho: NA
## Maximum Rhat value for EnvRho: 1.2038
## $n.chains
## [1] 5
##
## $n.adapt
## [1] 250000 250000 250000 250000 250000
##
## $sufficient.adapt
## [1] FALSE FALSE FALSE FALSE FALSE
##
## $n.iter
## [1] 20000
##
## $n.burnin
## [1] 0
##
## $n.thin
## [1] 100
##
## $n.samples
## [1] 1000
Gjam
data<-sim_data$CompSparseSp5
#fit_gjam(data,5000,2000,"./gjam_models/gjam5cmps.rda",interact= (-1)*comp_inter[[13]])
load_gjam(data,name="./gjam_models/gjam5cmps.rda", interact= (-1)*comp_inter[[13]])##
## Sensitivity by predictor variables f:
## Estimate SE CI_025 CI_975
## env 39.5 14.7 16.3 77.2
## env2 48.7 14.0 24.0 78.8
##
## Coefficient matrix B:
## sp01 sp02 sp03 sp04 sp05
## intercept -1.4600 -0.221 1.440 2.070 -0.115
## env -0.9350 -1.160 -0.284 3.070 2.070
## env2 0.0755 -1.220 -0.798 0.521 0.528
## RMSPE 0.3580 0.423 0.357 0.336 0.329
##
## Coefficient matrix B:
## Estimate SE CI_025 CI_975 sig95
## sp01_intercept -1.4600 0.0769 -1.6200 -1.320 *
## sp01_env -0.9350 0.0824 -1.1100 -0.781 *
## sp01_env2 0.0755 0.0610 -0.0379 0.197
## sp02_intercept -0.2210 0.0475 -0.3150 -0.125 *
## sp02_env -1.1600 0.0737 -1.3100 -1.020 *
## sp02_env2 -1.2200 0.0753 -1.3700 -1.080 *
## sp03_intercept 1.4400 0.0860 1.2700 1.610 *
## sp03_env -0.2840 0.0584 -0.3930 -0.165 *
## sp03_env2 -0.7980 0.0661 -0.9300 -0.669 *
## sp04_intercept 2.0700 0.3490 1.5300 2.690 *
## sp04_env 3.0700 0.3970 2.4300 3.740 *
## sp04_env2 0.5210 0.1090 0.3170 0.736 *
## sp05_intercept -0.1150 0.1890 -0.4600 0.291
## sp05_env 2.0700 0.1450 1.8000 2.350 *
## sp05_env2 0.5280 0.1950 0.1750 0.946 *
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Coefficient matrix B, standardized for X:
## NULL
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Coefficient matrix B, standardized for X and W:
## NULL
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Design Table
## env env2
## VIF 1 1
## env2 0 NA
##
## Sample contains n = 500 observations on S = 5 response variables, and 2 predictors. Data types (typeNames) include PA. There are 0 missing values in X and 0 missing values in Y. The RMSPE is 0.362, and the DIC is 45260. Computation involved 5000 Gibbs steps, with a burnin of 2000.
#gjfd5<-load_object("./gjam_models/gjam5cmps.rda")
#to check posterior density of s in Sigma
#gjfd5<-load_object("./gjam_models/gjam5cmps.rda")
#plot(density(gjfd5$chains$sgibbs[,4]))HMSC
data<-sim_data$CompSparseSp5
hm_mod<-fit_hmsc(data,"Load",nsamples=2000, nchains=2,name="./HMmodels/hm5cmps.rda" )
hm_conv(hm_mod)
hm_inter(hm_mod, nsamples=2000, nchains=2,interact = (-1)*comp_inter[[13]])
Environmental filtering + Competition sparse 10 species
Gjam
data<-sim_data$CompSparseSp10
#fit_gjam(data,2000,500,"./gjam_models/gjam10cmps.rda",interact= (-1)*comp_inter[[14]])
load_gjam(data,name="./gjam_models/gjam10cmps.rda", interact= (-1)*comp_inter[[14]])##
## Sensitivity by predictor variables f:
## Estimate SE CI_025 CI_975
## env 82 20.10 44.5 124.0
## env2 45 9.55 31.6 67.9
##
## Coefficient matrix B:
## sp01 sp02 sp03 sp04 sp05 sp06 sp07 sp08 sp09
## intercept -1.220 0.0536 0.645 0.406 1.420 1.110 0.0665 -0.1640 0.887
## env -1.670 -1.9900 -1.910 -0.238 -0.354 0.294 0.3120 1.0900 3.340
## env2 0.508 0.0778 -0.257 -0.998 -1.060 -0.581 -0.9340 -0.0165 0.105
## RMSPE 0.309 0.3470 0.341 0.436 0.354 0.393 0.4810 0.4600 0.277
## sp10
## intercept -0.023
## env 2.390
## env2 0.548
## RMSPE 0.315
##
## Coefficient matrix B:
## Estimate SE CI_025 CI_975 sig95
## sp01_intercept -1.2200 0.0857 -1.380000 -1.0400 *
## sp01_env -1.6700 0.1090 -1.870000 -1.4500 *
## sp01_env2 0.5080 0.0571 0.395000 0.6200 *
## sp02_intercept 0.0536 0.0694 -0.067400 0.1920
## sp02_env -1.9900 0.1090 -2.200000 -1.7700 *
## sp02_env2 0.0778 0.0677 -0.053200 0.2100
## sp03_intercept 0.6450 0.0544 0.538000 0.7490 *
## sp03_env -1.9100 0.1290 -2.180000 -1.6800 *
## sp03_env2 -0.2570 0.0484 -0.350000 -0.1630 *
## sp04_intercept 0.4060 0.0555 0.295000 0.5140 *
## sp04_env -0.2380 0.0728 -0.373000 -0.0994 *
## sp04_env2 -0.9980 0.0661 -1.130000 -0.8770 *
## sp05_intercept 1.4200 0.0719 1.280000 1.5600 *
## sp05_env -0.3540 0.0592 -0.477000 -0.2440 *
## sp05_env2 -1.0600 0.0625 -1.180000 -0.9430 *
## sp06_intercept 1.1100 0.0611 0.981000 1.2200 *
## sp06_env 0.2940 0.0505 0.194000 0.3920 *
## sp06_env2 -0.5810 0.0517 -0.681000 -0.4820 *
## sp07_intercept 0.0665 0.0581 -0.049200 0.1770
## sp07_env 0.3120 0.0594 0.193000 0.4190 *
## sp07_env2 -0.9340 0.0652 -1.060000 -0.8040 *
## sp08_intercept -0.1640 0.0490 -0.258000 -0.0667 *
## sp08_env 1.0900 0.0732 0.944000 1.2200 *
## sp08_env2 -0.0165 0.0553 -0.129000 0.0802
## sp09_intercept 0.8870 0.1040 0.705000 1.0800 *
## sp09_env 3.3400 0.1690 3.030000 3.6900 *
## sp09_env2 0.1050 0.0542 0.000574 0.2170 *
## sp10_intercept -0.0230 0.0613 -0.134000 0.1000
## sp10_env 2.3900 0.1260 2.160000 2.6500 *
## sp10_env2 0.5480 0.0675 0.405000 0.6680 *
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Coefficient matrix B, standardized for X:
## NULL
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Coefficient matrix B, standardized for X and W:
## NULL
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Design Table
## env env2
## VIF 1 1
## env2 0 NA
##
## Sample contains n = 500 observations on S = 10 response variables, and 2 predictors. Data types (typeNames) include PA. There are 0 missing values in X and 0 missing values in Y. The RMSPE is 0.377, and the DIC is 96021. Computation involved 2000 Gibbs steps, with a burnin of 500.
#to check posterior density of s in Sigma
#gjfd5<-load_object("./gjam_models/gjam10cmps.rda")
#plot(density(gjfd5$chains$sgibbs[,4]))data<-sim_data$CompSparseSp10
hm_mod<-fit_hmsc(data,"Load",nsamples=2000, nchains=2,name="./HMmodels/hm10cmps.rda" )
hm_conv(hm_mod)
hm_inter(hm_mod, nsamples=2000, nchains=2,interact = (-1)*comp_inter[[14]])
Environmental filtering + Competition sparse 20 species
Gjam
data<-sim_data$CompSparseSp20
#fit_gjam(data,2000,500,"./gjam_models/gjam20cmps.rda",interact= (-1)*comp_inter[[15]])
load_gjam(data,name="./gjam_models/gjam20cmps.rda", interact= (-1)*comp_inter[[15]])##
## Sensitivity by predictor variables f:
## Estimate SE CI_025 CI_975
## env 288 84.9 156 457
## env2 128 21.3 94 175
##
## Coefficient matrix B:
## sp01 sp02 sp03 sp04 sp05 sp06 sp07 sp08 sp09
## intercept -0.366 -0.677 -0.599 0.151 0.2260 0.817 0.510 0.979 1.070
## env -1.990 -1.940 -2.400 -1.920 -1.7600 -1.760 -0.949 -1.030 -0.562
## env2 0.777 -0.456 -0.564 -0.531 0.0776 -0.247 -0.489 -0.535 -0.856
## RMSPE 0.316 0.361 0.321 0.329 0.3680 0.346 0.423 0.364 0.386
## sp10 sp11 sp12 sp13 sp14 sp15 sp16 sp17
## intercept 1.020 0.8200 1.180 1.050 1.090 0.757 0.7880 0.479
## env -0.291 0.0736 0.465 0.842 1.330 1.520 1.8300 2.380
## env2 -0.928 -0.6110 -0.634 -0.547 -0.554 -0.353 -0.0702 -0.104
## RMSPE 0.421 0.4590 0.383 0.383 0.330 0.356 0.3450 0.317
## sp18 sp19 sp20
## intercept 0.0281 -0.2080 -0.808
## env 1.9000 2.0400 2.280
## env2 -0.0839 0.0373 -0.101
## RMSPE 0.3490 0.3430 0.328
##
## Coefficient matrix B:
## Estimate SE CI_025 CI_975 sig95
## sp01_intercept -0.3660 0.0984 -0.5770 -0.20900 *
## sp01_env -1.9900 0.1510 -2.2600 -1.71000 *
## sp01_env2 0.7770 0.0681 0.6550 0.91500 *
## sp02_intercept -0.6770 0.0865 -0.8320 -0.49100 *
## sp02_env -1.9400 0.0850 -2.1100 -1.78000 *
## sp02_env2 -0.4560 0.0532 -0.5570 -0.35000 *
## sp03_intercept -0.5990 0.0619 -0.7310 -0.48600 *
## sp03_env -2.4000 0.1150 -2.6200 -2.18000 *
## sp03_env2 -0.5640 0.0586 -0.6800 -0.45600 *
## sp04_intercept 0.1510 0.0661 0.0233 0.27700 *
## sp04_env -1.9200 0.1190 -2.1200 -1.67000 *
## sp04_env2 -0.5310 0.0597 -0.6460 -0.41500 *
## sp05_intercept 0.2260 0.0559 0.1200 0.33300 *
## sp05_env -1.7600 0.0851 -1.9400 -1.60000 *
## sp05_env2 0.0776 0.0593 -0.0411 0.19100
## sp06_intercept 0.8170 0.0581 0.6950 0.92700 *
## sp06_env -1.7600 0.0971 -1.9500 -1.58000 *
## sp06_env2 -0.2470 0.0566 -0.3600 -0.14000 *
## sp07_intercept 0.5100 0.0540 0.4030 0.61700 *
## sp07_env -0.9490 0.0571 -1.0600 -0.83000 *
## sp07_env2 -0.4890 0.0590 -0.6070 -0.37800 *
## sp08_intercept 0.9790 0.0615 0.8600 1.10000 *
## sp08_env -1.0300 0.0912 -1.1900 -0.85900 *
## sp08_env2 -0.5350 0.0543 -0.6450 -0.42800 *
## sp09_intercept 1.0700 0.0667 0.9430 1.21000 *
## sp09_env -0.5620 0.0685 -0.6940 -0.43000 *
## sp09_env2 -0.8560 0.0573 -0.9660 -0.74400 *
## sp10_intercept 1.0200 0.0626 0.9000 1.15000 *
## sp10_env -0.2910 0.0502 -0.3850 -0.19200 *
## sp10_env2 -0.9280 0.0578 -1.0400 -0.81900 *
## sp11_intercept 0.8200 0.0553 0.7100 0.93000 *
## sp11_env 0.0736 0.0500 -0.0251 0.16800
## sp11_env2 -0.6110 0.0570 -0.7290 -0.50000 *
## sp12_intercept 1.1800 0.0596 1.0600 1.30000 *
## sp12_env 0.4650 0.0592 0.3480 0.58300 *
## sp12_env2 -0.6340 0.0529 -0.7400 -0.53500 *
## sp13_intercept 1.0500 0.0582 0.9290 1.16000 *
## sp13_env 0.8420 0.0529 0.7340 0.93700 *
## sp13_env2 -0.5470 0.0493 -0.6370 -0.44800 *
## sp14_intercept 1.0900 0.0690 0.9610 1.22000 *
## sp14_env 1.3300 0.0651 1.2000 1.46000 *
## sp14_env2 -0.5540 0.0574 -0.6730 -0.44200 *
## sp15_intercept 0.7570 0.0580 0.6440 0.86900 *
## sp15_env 1.5200 0.0708 1.3800 1.66000 *
## sp15_env2 -0.3530 0.0507 -0.4540 -0.25600 *
## sp16_intercept 0.7880 0.0591 0.6680 0.89500 *
## sp16_env 1.8300 0.1020 1.6300 2.02000 *
## sp16_env2 -0.0702 0.0457 -0.1550 0.02360
## sp17_intercept 0.4790 0.0492 0.3830 0.56900 *
## sp17_env 2.3800 0.0979 2.1900 2.58000 *
## sp17_env2 -0.1040 0.0494 -0.2020 -0.00328 *
## sp18_intercept 0.0281 0.0502 -0.0737 0.12500
## sp18_env 1.9000 0.1020 1.7100 2.10000 *
## sp18_env2 -0.0839 0.0534 -0.1840 0.02740
## sp19_intercept -0.2080 0.0626 -0.3440 -0.09310 *
## sp19_env 2.0400 0.1380 1.7600 2.29000 *
## sp19_env2 0.0373 0.0524 -0.0618 0.13600
## sp20_intercept -0.8080 0.0647 -0.9280 -0.67900 *
## sp20_env 2.2800 0.1060 2.0700 2.48000 *
## sp20_env2 -0.1010 0.0547 -0.2090 0.00857
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Coefficient matrix B, standardized for X:
## NULL
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Coefficient matrix B, standardized for X and W:
## NULL
##
## Last column indicates if 95% posterior distribution contains zero.
##
## Design Table
## env env2
## VIF 1 1
## env2 0 NA
##
## Sample contains n = 500 observations on S = 20 response variables, and 2 predictors. Data types (typeNames) include PA. There are 0 missing values in X and 0 missing values in Y. The RMSPE is 0.363, and the DIC is 340096. Computation involved 2000 Gibbs steps, with a burnin of 500.
#to check posterior density of s in Sigma
#gjfd5<-load_object("./gjam_models/gjam20cmps.rda")
#plot(density(gjfd5$chains$sgibbs[,4]))HMSC
data<-sim_data$CompSparseSp20
hm_mod<-fit_hmsc(data,"Load",nsamples=2000, nchains=2,name="./HMmodels/hm20cmps.rda" )
hm_conv(hm_mod)
hm_inter(hm_mod, nsamples=2000, nchains=2,interact = (-1)*comp_inter[[15]])
Environmental filtering + Competition +Facilitation 5 dense species
Environmental filtering + Competition +Facilitation 10 dense species
Environmental filtering + Competition +Facilitation 20 dense species
Graph from paper
Visualization for true interactions
